This diversity action arises out of the fatal crash of a General Dynamics F-16 fighter aircraft piloted by Air Force Captain Theodore Harduvel. After a two-week trial, a jury found General Dynamics liable for defects in the plane and awarded Harduvel's estate and survivors a total of $3.1 million in damages. The primary questions presented in this appeal concern the sufficiency of the evidence in support of this verdict and General Dynamics' entitlement to the federal common law government contractor's defense. Because we think the award of damages in this case is inconsistent with the Supreme Court's recent formulation of the government contractor defense in Boyle v. United Technologies Corp., --- U.S. ----, 108 S. Ct. 2510, 101 L. Ed. 2d 442 (1988), we reverse.

The subject of this products liability suit is the General Dynamics F-16 "Fighting Falcon" aircraft. The F-16 was developed to replace the F-4 "Phantom" as a primary fighter aircraft for the United States Air Force. The F-16 is the first fighter that employs a "fly-by-wire" system, designed at the request of the government. All of the controls in the aircraft are electric. The control stick is not, as in other aircraft, connected to control surfaces by levers and wires. Rather, the pilot's hand pressure on the stick controls the craft through electric impulses. The F-16 entered operational service with the Air Force in 1979. The Air Force has acquired approximately 1,700 of the planes, including the the original F-16A, two-seat F-16B, and later-model F-16C and F-16D, and plans a total force of approximately 2,700. A number of foreign governments have purchased the F-16 for use in their air forces. Some F-16's have been manufactured abroad through cooperative ventures among foreign governments, the United States, and General Dynamics. Captain Harduvel's aircraft was an F-16A, serial number 81-692, manufactured by General Dynamics at its Fort Worth, Texas plant and accepted by the Air Force in August 1982.

Captain Harduvel's tragic crash occurred on November 15, 1982, near Kunsan Air Base in South Korea. Harduvel, an experienced military pilot and flight instructor, was leading a flight of three F-16A fighter aircraft on a routine training mission. The other two pilots accompanying Harduvel were Captain David L. Moody, the subject of the training exercise, and Captain Andy C. Denny, who flew behind the other planes in a chase position. The planes were flying at an altitude of approximately 5,000 feet and an airspeed of 480 knots. About eight minutes into the flight, Harduvel radioed the other pilots, "Knock it off. One has a problem. Two join on me. Three continue the route." In military vernacular, this informed the other pilots that Harduvel was experiencing unspecified difficulty, ordered Moody to maneuver closer to Harduvel, and ordered Denny to continue the mission.

After sending this radio message, Harduvel began a hard left-hand turn back toward the direction of the airbase. Before Moody could join up with Harduvel's craft, Harduvel left available clear air and flew into a small cloud. He emerged from this cloud, and then entered a solid cloud bank. Both Moody and Denny testified that at the time they last saw Harduvel's airplane, it appeared to be flying normally. Although there were no witnesses to the final seconds of the flight, both parties appear to accept that Harduvel for some reason became disoriented in the clouds and inadvertently allowed the plane to go into a steep dive. Shortly after turning into the cloud, Harduvel's aircraft hit the back of a mountain ridge at a downward angle of approximately sixty degrees, with wings level. Captain Harduvel was killed and the aircraft almost completely destroyed.

Captain Harduvel's widow, Janet Harduvel, filed this diversity action against General Dynamics in her home state of Florida in January 1985. She sued individually and on behalf of Harduvel's estate and her minor child, Christina Harduvel. The case was tried before a jury in April 1987. At trial, plaintiff and defendant advanced diametrically opposed theories of the reasons for Captain Harduvel's turn into the clouds and ensuing crash. There were no eye witnesses to what happened in the clouds. Both parties therefore relied on witnesses who qualified as experts. As often happens in this situation, the experts disagreed.

Plaintiff's experts expressed their opinion that Harduvel's F-16 suffered a massive electrical failure that caused all of his flight instruments to malfunction. Without flight instruments, he became disoriented in the clouds. The witnesses theorized that "wire chafing"--the rubbing of wires in the electrical system against other wires, fasteners, or structural parts of the plane--caused electrical shorts, overheating, and fire in the "right strake" of the aircraft, a forward extension of the wing located beneath and to the side of the cockpit. According to this theory, the electrical malfunction caused a loss of power to the plane's attitude indicators, the instruments that tell the pilot the position of his wings and whether the nose of the aircraft is pointed up or down.

Plaintiff also produced expert testimony to the effect that the Inertial Navigation Unit (INU), that provides primary attitude reference information, was not receiving power at the time of the crash. Other testimony theorized that the Standby Attitude Indicator (SAI), that provides a backup attitude reference, was displaying an incorrect reading at impact. Plaintiff's experts also testified to a history of wire chafing in F-16 airplanes, and contended that the design of the fighter's electrical system would allow chafing to cause electrical faults that could spread throughout the system and result in a loss of power to flight instruments.

Defendant, also relying on the opinions of expert witnesses, advanced a different theory of the crash. It contended that Harduvel suffered a severe reaction to medication that he had been taking for a prostate infection. Harduvel had been grounded for two weeks due to this condition, and the fateful flight was his first after being certified fit to resume flying. Defendant contended that the medication, Bactrim, caused Harduvel to experience severe nausea and discomfort during the flight. This prompted Harduvel to abort the mission and attempt to return to Kunsan, and caused him to lose control of the aircraft in the clouds. Defendant also placed great emphasis on the undisputed fact that Harduvel flew into clouds rather than remain in clear air where visibility was good. In defendant's view, whether the problem facing Harduvel was physical or electrical, the turn into the clouds was fatal pilot error, and the ultimate cause of the accident.

Defendant also produced several expert witnesses on the electrical system of the F-16, who testified in support of the theory that wire chafing was not involved in Harduvel's crash. Several of these experts testified as a fact that there had never been a report of an F-16 crash attributed to wire chafing and never been a report of a failure of flight instruments due to chafing. They testified that the F-16 employs redundant wiring to prevent system failures. Defendant's witnesses also gave opinion testimony that the flight control systems on Harduvel's craft were receiving power and functioning normally at the time of the crash. One of these witnesses speculated that the incorrect reading on the SAI when it was found resulted from the force of the crash, and that the instrument was operating correctly at impact.

The jury was instructed on both negligence and strict liability as to plaintiff's claims of design defect, manufacturing defect, and failure to warn. Answering special interrogatories, the jury found in plaintiff's favor on each of the six claims: that defendant was negligent in designing the F-16, negligent in its manufacture, negligent in failing to warn, strictly liable for design and manufacturing defects in the aircraft, and strictly liable for failure to give adequate warnings. The jury also found against defendant on its defenses of comparative negligence and the federal government contractor's defense. The jury awarded damages of $1.6 million to Captain Harduvel's estate, and $750,000 each to Janet and Christina Harduvel.

Defendant moved for JNOV on all claims, and the district court issued a written opinion. The court denied defendant's JNOV motions as to the design and manufacturing defect claims, thus leaving intact the $3.1 million award, but set aside the verdict on failure to warn, finding that there had not been sufficient proof to take that issue to the jury. The district court also denied a motion by the plaintiff for prejudgment interest. In its opinion, the district court noted that the defendant did not challenge the jury verdicts on the defenses in the JNOV motion, but observed that "the military contractor defense was unsuccessful under a correct application of the test set forth by the Eleventh Circuit in Shaw v. Grumman Aerospace Corp., 778 F.2d 736, 745-6 (11th Cir.1985)."

Both parties appealed. Plaintiff challenged the district court's entry of JNOV on the failure to warn claim and its denial of plaintiff's motion for prejudgment interest. Defendant argued that the district court erred in denying its motions for JNOV and for a new trial on the manufacturing and design defect verdicts. The defendant relied specifically on the government contractor's defense.

Nine months after the district court issued its memorandum decision in this case, the Supreme Court decided Boyle v. United Technologies Corp., --- U.S. ----, 108 S. Ct. 2510, 101 L. Ed. 2d 442 (1988). In Boyle, the Court recognized a broad formulation of the government contractor defense that shields manufacturers of products made for the government from tort liability for flaws in product design. The Court recognized that in certain areas of "uniquely federal interests," state law must be preempted, and if necessary replaced, by federal common law. One such area of uniquely federal interest is the government's procurement of military hardware. The Court grounded the contours of the defense in the "discretionary function" exception to the Federal Tort Claims Act, 28 U.S.C. Sec. 2680(a), that protects the United States from liability for its agents' performance of duties involving discretionary decisions. 108 S. Ct. at 2514-2518. Without the defense, the government's own tort immunity for its discretionary functions would be undermined. Contractors held liable for design features that were the subject of discretionary approval by the government would predictably pass on the costs of liability, ultimately imposing costs on the government that its immunity was intended to preclude. See id. at 2518.

The defense derives from the principle that where a contractor acts under the authority and direction of the United States, it shares the sovereign immunity that is enjoyed by the government. See Yearsley v. W.A. Ross Construction Co., 309 U.S. 18, 60 S. Ct. 413, 84 L. Ed. 554 (1940). In the military context, this immunity serves the further important purpose of shielding sensitive military decisions from scrutiny by the judiciary, the branch of government least competent to review them. Application of ordinary tort law to military design and procurement decisions is not appropriate, for the government "is required by the exigencies of our defense effort to push technology towards its limits and thereby incur risks beyond those that would be acceptable for ordinary consumer goods." Tozer v. LTV Corp., 792 F.2d 403, 406 (4th Cir.1986) (quoting McKay v. Rockwell International Corp., 704 F.2d 444, 449-50 (9th Cir.1983)).1

Applying these principles, the Court based its formulation of the defense on the earlier decisions of the Fourth and Ninth Circuits in Tozer and McKay, supra:

Liability for design defects in military equipment cannot be imposed, pursuant to state law, when (1) the United States approved reasonably precise specifications; (2) the equipment conformed to those specifications; and (3) the supplier warned the United States about the dangers in the use of the equipment that were known to the supplier but not to the United States.

108 S.Ct. at 2518. The three conditions serve to ensure that the defense operates to immunize the contractor only where the government has actually participated in discretionary design decisions, either by designing a product itself or approving specifications prepared by the contractor. See id. at 2517.

Defendant sought summary judgment in this case under the government contractor defense, and at trial asked for an instruction incorporating the Tozer and McKay standard later adopted by the Supreme Court. The district court denied the summary judgment motion, and instructed the jury on the basis of this circuit's formulation of the defense in Shaw, noting the distinction between Shaw and the test used in other circuits. See R4-148, Rl6-12, Rl8-195-196. Under Shaw, the contractor is entitled to the defense

only if it affirmatively proves: (1) that it did not participate, or participated only minimally, in the design of those products or parts of products shown to be defective; or (2) that it timely warned the military of the risks of the design and notified it of alternative designs reasonably known by the contractor, and that the military, although forewarned, clearly authorized the contractor to proceed with the dangerous design.

778 F.2d at 746. In Boyle, the Supreme Court specifically rejected the Shaw formulation of the government contractor defense. The Court stated that

it is not a rule designed to protect the federal interest embodied in the "discretionary function" exemption. The design ultimately selected may well reflect a significant policy judgment by Government officials whether or not the contractor rather than those officials developed the design. In addition, it does not seem to us sound policy to penalize, and thus deter, active contractor participation in the design process, placing the contractor at risk unless it identifies all design defects.

108 S.Ct. at 2518. The primary question before us now is the effect of the intervening Boyle decision on plaintiff's damage award.

Before turning to the application of the government contractor defense, we must first address plaintiff's contention that the jury's finding of a manufacturing defect makes the contractor defense inapplicable here. Plaintiff correctly points out that Boyle, by its terms, applies only to defects in design. See id.; see also McGonigal v. Gearhart Industries, Inc., 851 F.2d 774, 777 (5th Cir.1988). The defense adopted in Boyle "does not relieve suppliers of military equipment of liability for defects in the manufacture of that equipment." McKay, 704 F.2d at 451. We think plaintiff's case is subject to the government contractor defense notwithstanding the manufacturing defect verdict for two reasons. First, although plaintiff would describe her theory of wire chafing and electrical failure as one involving defective manufacture, we think it clearly is a claim of defective design for purposes of the government contractor defense. Second, the specific evidence on which the district court relied to justify its denial of JNOV on the manufacturing defect claims does not suffice to support the jury's verdict.

In light of the Boyle decision, the distinction between manufacturing and design defect claims takes on special significance in the context of accidents such as this one. Plaintiff contends that a finding of defective manufacture under state law suffices to remove the government contractor defense from the case. She also relies heavily on the fact that some witnesses in the case, including General Dynamics' own experts, referred to wire chafing as a defect of "manufacture." Rl3-205-206. But whether a defect is properly designated as one of design or manufacture is not governed by a witness' choice of terminology, or indeed by state tort law. The government contractor defense is a matter of federal common law, and so is the denomination of a defect as one of design or manufacture for purposes of applying the defense. Were this not so, state law could operate either to defeat the defense or to expand it improperly, and the defense could not be applied with the uniformity that is a key justification for application of federal common law. See, e.g., Clearfield Trust Co. v. United States, 318 U.S. 363, 367, 63 S. Ct. 573, 575, 87 L. Ed. 838 (1943).

For this reason, we cannot accept plaintiff's initial contention that the government contractor defense is inapplicable here due to a Florida law presumption of manufacturing defect in cases where an allegedly defective product is destroyed. Under Florida law, where a product is destroyed in an accident, and the plaintiff presents evidence to negate possible causes other than a product defect, an inference of manufacturing defect arises. See Worsham v. A.H. Robins Co., 734 F.2d 676 (11th Cir.1984); Cassisi v. Maytag Co., 396 So. 2d 1140 (Fla.D.Ct.App.1981). Although the district court declined to apply the inference here for other reasons, we think the manufacturing defect inference is incompatible with the government contractor defense. Military equipment is highly likely to be destroyed in accidents such as Harduvel's. Thus, the Florida law inference would undermine the Boyle defense by creating a presumption that it is inapplicable. In cases implicating the federal interest recognized in Boyle, therefore, the inference of manufacturing defect is displaced by federal common law.

In determining whether a claimed defect is one of "design" for purposes of the government contractor defense, the proper focus is the protection of discretionary government functions for which the defense is intended. If a defect is one inherent in the product or system that the government has approved, it will be covered by the defense. Where a defect is merely an instance of shoddy workmanship, it implicates no federal interest. This distinction between "aberrational" defects and defects occurring throughout an entire line of products is frequently used in tort law to separate defects of manufacture from those of design. See, e.g., M. Franklin & R. Rabin, Cases and Materials on Tort Law and Alternatives 497 (4th ed.1987). Stated another way, the distinction is between an unintended configuration, and an intended configuration that may produce unintended and unwanted results.

Applying these principles to the record before us, we think plaintiff's theory of the fatal crash is based on a claim of generally defective design in the F-16 electrical system. The expert on whom the district court primarily relied, Robert Alexander, made numerous reference to "product-wide" defects in the electrical system, and failure to follow proper design procedures to avoid the possibility of chafing. See, e.g., Rl2-53, 12-89, 12-93, 12-98, 12-135, 12-140, 13-51. Alexander based many of his opinions on plaintiff's Exhibit 48-D, a compilation of field reports of wire chafing prepared by the Air Force. Rl2-159. Review of Exhibit 48-D and Alexander's testimony reveals that he based his opinion that wire chafing occurred in Harduvel's craft on chafing reports involving many different components of many different aircraft, Rl2-159, and not on evidence of a particular defect in the way Harduvel's plane was manufactured.

Plaintiff's Exhibit 122, a training videotape about wire chafing produced for Air Force maintenance personnel by General Dynamics under contract with the Air Force, also illustrates the nature of plaintiff's claim.2 The film's narrator states:

... changes in maintenance procedures go hand in hand with advanced aerospace technology. The F-16 has more than eleven miles of wire inside its air-frame, all of which is subjected to constant vibration, as well as forces up to nine G's. The forces cause movement of the wire harnesses within the structure of the aircraft. This movement can cause chafing.

This film, on which plaintiff has placed heavy reliance throughout this litigation, indicates that wire chafing is a problem inherent in the design of the all-electric F-16.

In sum, the chafing problem as presented in this case is one of placing many wires close together without adequate insulation. Rl2-53-54, 13-64. The government's promulgation of Military Specification 5088E, a general proscription of chafing that applied to all aircraft, as well as the trial testimony of electrical experts, Rl6-205, shows that wire chafing was a problem with aircraft electrical systems that had long been known both in the industry and to the Air Force. The Air Force's decision to produce the first "all-electric" fighter could thus be expected to bring increased risks of chafing. Moreover, with full knowledge of the chafing problem in the design of the aircraft, the Air Force has continued since 1979 to purchase the F-16 as a primary combat fighter plane. Plaintiff has depicted Harduvel's plane not as an aberrational example of improper manufacture, but as a typical F-16 with a design that includes a potential for unwanted wire chafing.

Of course, the fact that plaintiff's general theory of the case presented wire chafing as a product-wide design flaw would not preclude proof of a manufacturing defect as well. But our review of the specific evidence on which plaintiff relies in defense of the manufacturing defect verdict convinces us that the district court erred in denying defendant's JNOV motion.

The wiring system of Harduvel's aircraft was completely destroyed in the crash, and the only evidence concerning wire chafing in this plane was testimony from expert Alexander, who examined the fighter's maintenance records. These records did not report that chafed wires had been found in the aircraft. Rather, Alexander focused on the fact that several of the maintenance items involved in-flight malfunctions that later could not be duplicated on the ground. Rl2-90-98. He testified that malfunctions that occur in flight but cannot later be duplicated by maintenance personnel are often associated with wire chafing. This evidence is not sufficient to show the presence of a manufacturing defect. Alexander did not identify any particular site of chafing on Harduvel's F-16, and gave no opinion as to the origin of chafing shown to have existed. His testimony only supported plaintiff's theory that F-16's generally were susceptible to some wire chafing. Rl2-89.

Plaintiff also contends that Harduvel's plane was defectively manufactured in that it contained an improper DC power panel. Plaintiff's Exhibit 67, "Engineering Change Proposal 712" describes a change that was made to a DC power panel in the right strake area of F-16 fighters. The height of the panel was reduced in order to alleviate chafing that was "primarily a result of wiring density and a tight fit" in the area. The new panel was not installed on Harduvel's aircraft, and plaintiff contends this shows defective manufacture. Examination of Exhibit 67, however, reveals that the change in the power panel took effect only with production Block 15E, while Harduvel's plane came from the earlier Block 15C. Exhibit 67 also indicates that the Air Force issued waivers covering aircraft manufactured before the change. The DC power panel in Harduvel's plane was thus installed as intended by the existing design and accepted by the Air Force. To the extent this was a defect, it was not one of manufacture, and cannot support the manufacturing defect verdict.

Plaintiff, as did the district court below, relies most heavily for support of the manufacturing defect claim on a maintenance report found at pages 229-230 of Plaintiff's Exhibit 48-D.3 This report describes an incident in which a center screw on panel 2206 in the right strake of an F-16 "protruded" into a wire harness bundle, causing chafing damage. The report concluded that the chafing was produced when maintenance caused the wire harness to be placed in contact with the screw. The report goes on to state that "the contractor has inspected seven aircraft in final assembly at this facility and found the same potential problem exists." The report does not state whether chafing had occurred in the seven aircraft, or whether the "potential problem" was that wires might be moved into contact with screws in the course of future maintenance. The date of the report was September 1982. Harduvel's plane was accepted by the Air Force the previous month. Alexander testified on this basis that Harduvel's plane probably had the same defect, and that this defect was the probable cause of the crash. Rl2-167-169.

We think the district court erred in holding that the exhibit and testimony sufficed to support the manufacturing defect verdict. First, we cannot accept plaintiff's repeated characterization of the report as one concerning "improperly installed screws." No record evidence supports the proposition that any screw on the F-16 that "protruded" must have been improperly installed. Indeed, the opposite is true. Plaintiff's Exhibit 122, the maintenance training film on wire chafing, see supra at 14, describes wire contact with protruding screws as a common occurrence, and depicts protruding screws in a normal F-16. Similarly, Air Force Time Compliance Technical Order (TCTO) 1078, a September 1982 document outlining inspection procedures for the prevention of wire chafing, directed maintenance personnel to "inspect all ... protruding screws, hardware, and other objects." Plaintiff's Ex. 48-E. The record indicates that protruding screws were a normal condition in the F-16, not a product of improper installation. The record also shows that the screws involved in the potential chafing described in exhibit 48-D were standard Air Force equipment that defendant was required to use. Rl7-48-49. The record does not show that the placement or use of the screws conflicted with any design specifications.

Moreover, only seven F-16's in defendant's plant were shown to have a potential chafing problem caused by the protruding center screw. These seven had not left the factory, and had not yet been accepted by the Air Force. Rl2-170. There is no evidence that any F-16 entered active service with the particular chafing problem described in exhibit 48-D, and no direct evidence that this type of wire chafing ever occurred on Harduvel's craft. Plaintiff cannot rely on a sampling of seven aircraft out of a production block of 300, Rl6-40, to establish that this particular defect was present in Harduvel's plane and caused the crash. Cf. Nichols Construction Corp. v. Cessna Aircraft Co., 808 F.2d 340, 347-52 (5th Cir.1985) (evidence showing that 27 out of 15,000 "retention rings" manufactured were cracked insufficient to support finding that defective ring caused a particular crash).

Reviewing the evidence in the light most favorable to plaintiff, we do not think there is "evidence of such quality and weight that reasonable and fair-minded men in the exercise of impartial judgment," Boeing Co. v. Shipman, 411 F.2d 365, 375 (5th Cir.1969) (en banc), could conclude that plaintiff proved the presence of a manufacturing defect on Harduvel's F-16, and we hold that the district court erred in denying defendant's motion for JNOV on the manufacturing defect claims.

We therefore apply the three conditions for the government contractor defense set forth in Boyle to the record before us. The first requires that the government "approved reasonably precise specifications." 108 S. Ct. at 2518. We think the F-16 design process meets this condition as a matter of law. The Air Force initiated the F-16 project in the early 1970's by issuing a Request for Proposal seeking a contractor to develop the first "fly-by-wire" or all-electric fighter aircraft. Rl7-126. General Dynamics responded with proposed Air Vehicle Specifications, that were evaluated by the Air Force Systems Program Office. The Office conducted an extensive review of the aircraft, including the electrical system, examining specifications, drawings, and blueprints. Rl7-106, 130, 142. One group of Air Force engineers was specifically assigned to review of the electrical system. Rl7-144. The Air Force conducted independent review and analysis of the electrical system design, Rl5-186, 17-129, and evaluated designs in a Preliminary Design Review, Critical Design Review, and later Physical Configuration Audit that took place between 1973 and 1975 prior to the start of production. Rl7-23-24, 17-110. Government review and approval of design and production methods continued after production began. Indeed, the Air Force requested an increase in wiring capacity on the particular block of F-16's from which Harduvel's came. Rl6-34, 16-106-107. The defense requires "only that the government approve reasonably precise specifications," and is met where, as here, the contractor incorporated government performance specifications into a design that the government subsequently reviewed and approved. Smith v. Xerox Corp., 866 F.2d 135, 138 (5th Cir.1989).

In sum, the design of the F-16 was a result of "continuous back and forth" between the Air Force and General Dynamics. Koutsoubos v. Boeing Vertol, 755 F.2d 352, 355 (3d Cir.), cert. denied 474 U.S. 821, 106 S. Ct. 72, 88 L. Ed. 2d 59 (1985). Although one court has held that the contractor may not avail itself of the defense when the government's approval is nothing more than a "rubber stamp," see Trevino v. General Dynamics Corp., 865 F.2d 1474, 1480 (5th Cir.1989), that is not the case here. The district court recognized as much in denying defendant's motion for summary judgment under the now-rejected Shaw standard, stating that "the Air Force approved the design specifications submitted by General Dynamics," and that "there was a great deal of communication between the government and General Dynamics concerning the aircraft." R4-128. Air Force acquisition of the F-16 was not, as plaintiff contends, a mere purchase of stock aircraft. Cf. Boyle, 108 S. Ct. at 2516. Rather, the F-16 was designed through the cooperative "back-and-forth" that is the "reality of the procurement process, as well as a valuable part of that process." Tozer, 792 F.2d at 407.

The second condition for the defense requires that the equipment conform to the specifications approved. 108 S. Ct. at 2518. This prong of Boyle involves essentially the same analysis that we have discussed above with respect to the distinction between manufacturing and design defects. See Zinck v. ITT Corp., 690 F. Supp. 1331, 1337 (S.D.N.Y.1988). To say that a product failed to conform to specifications is just another way of saying that it was defectively manufactured. Indeed, plaintiff does not independently address the second prong of Boyle in her briefs. For the reasons discussed above, we do not find substantial evidence in the record to support plaintiff's assertion that a manufacturing defect in the electrical system of Harduvel's F-16 was a cause of the crash. Plaintiff did present substantial evidence that the design--not the manufacture--of the F-16 wiring system was susceptible to dangerous wire chafing, but that claim is subject to the Boyle defense. The second condition for application of the defense is present here as a matter of law.

The final Boyle condition requires that the supplier warn the United States about dangers in the use of the equipment that were known to the supplier but not to the United States. 108 S. Ct. at 2518. With respect to this claim, we note that the district court granted defendant's JNOV motion as to the jury's finding that defendant was liable for failure to warn. We agree with the district court's determination that there was insufficient evidence to place this claim before the jury. The record shows that the Air Force knew of wire chafing by 1979 or 1980, well before Harduvel's crash in 1982. Rl6-27, 17-56. Defendant produced uncontested evidence that its engineers withheld no information on chafing or other problems from the Air Force. Rl6-26, 16-195. Indeed, much of General Dynamics' knowledge of wire chafing came from Air Force field reports. Rl6-27. Plaintiff produced no example of information about the chafing problem that was known to General Dynamics but not conveyed to the Government.

Plaintiff goes a step further, and argues that even if General Dynamics shared knowledge of chafing, it failed to inform the Air Force of the "seriousness" of the chafing problem, causing the Air Force to take insufficient preventive measures. Plaintiff relies on Exhibit 122, the maintenance training film, which states that uncorrected chafing problems can lead to "disastrous" consequences. Because the film was released to the Air Force only after Harduvel's crash, plaintiff contends that General Dynamics at that time was withholding knowledge of the true seriousness of the chafing problem, specifically that chafing could lead to a fatal loss of aircraft. We cannot agree, for there is no evidence in the record that General Dynamics possessed knowledge that chafing was a graver danger than was known by the Air Force. We repeat that the record contains no evidence that a fatal crash had ever been attributed to wire chafing. General Dynamics presented uncontradicted evidence that there had been no reports of a crash caused by chafing, Rl6-65, 16-183, 16-209, 17-37, no reports of navigation instrument failure due to chafing, Rl6-65, and no reports of fire in the right strake area of an F-16, Rl6-209.

Air Force technical orders concerning inspections for chafing, contained in Plaintiff's Exhibit 48-E, reveal that the government was aware of the seriousness of the problem. For example, TCTO-1078, in which the Air Force mandated inspections for chafing at 100-hour intervals, listed numerous areas that must be inspected, including essential aircraft systems and the very right strake panels on which plaintiff's evidence focused. The order stated that "[f]ailure to accomplish this inspection may result in loss of one or more systems caused by a chafed harness." This order was issued in September 1982, over two months prior to Harduvel's crash. Technical order 1F-16A-508, dated March 22, 1982, warned that chafing could lead to electrical shorting. Rl6-104. Other technical orders in Exhibit 48-E similarly warn that failure to accomplish them may result in electrical shorts and loss of aircraft systems.

The Air Force was plainly aware of the chafing problem in the F-16 generally, and of the possible serious consequences of chafing. There is no evidence that General Dynamics had knowledge it withheld. Despite its knowledge, the Air Force continued to fly the hundreds of F-16's in operation, and to purchase additional ones. Its decision was simply to address the chafing problem through the regular 100-hour inspections and technical orders. Rl6-49. We think the third Boyle condition is satisfied here as a matter of law.

Young servicemen such as Captain Theodore Harduvel represent the very best of our Nation's citizens. Americans take pride in their bravery and skill, and mourn when their lives are tragically lost. The pilots and crews of military aircraft willingly embrace the risks that they assume by volunteering to serve our country. They are not the "military doubles of civilian motorists," Tozer, 792 F.2d at 407, or ordinary purchasers of consumer products. The Supreme Court's adoption of the government contractor defense recognizes that one of these risks is the operation of equipment in which safety concerns have been balanced against cost and performance. With respect to consumer goods, state tort law may hold manufacturers liable where such a balance is found unreasonable. In the sensitive area of federal military procurement, however, the balance is not one for state tort law to strike. Although the defense may sometimes seem harsh in its operation, it is a necessary consequence of the incompatibility of modern products liability law and the exigencies of national defense.

We conclude that plaintiff's claims are subject to the government contractor defense, and that the record before us establishes the conditions of the defense. Accordingly, we reverse the judgment of the district court and remand for entry of judgment in favor of General Dynamics.

REVERSED.