This appeal in a personal-injury diversity suit brings up a variety of interesting substantive and procedural questions. In 1978 a middle-aged man named Forrest Flaminio bought a "Gold Wing" motorcycle, a large and powerful touring motorcycle manufactured by the Honda Motor Company of Japan (Japanese Honda) and distributed in the United States by its wholly owned subsidiary, American Honda Motor Company (American Honda). The motorcycles are shipped from Japan partially assembled, but assembly is completed by the dealers to whom American Honda distributes the motorcycles rather than by American Honda itself. Three days after taking delivery, and shortly after a dinner at which he had one or two drinks, Flaminio was driving the motorcycle down a two-lane road at night when he came up behind a car traveling at about 40 miles per hour. He passed it at a speed of somewhere between 50 and 70 m.p.h. (the speed limit was 50), and as he did so felt a vibration in the front end of the motorcycle. He tried to look at the front wheel to see what was wrong. This was an awkward maneuver because his feet were up on the motorcycle's "highway pegs" (supplied and installed by the motorcycle dealer rather than by either of the defendants), so that he was leaning backward. By his own admission the effort in this position to see the front wheel probably brought him up off the seat. In any event the motorcycle began to wobble uncontrollably and then it shot off the road and crashed, inflicting injuries that have left Flaminio a paraplegic.

Joined by his wife, who is seeking to recover damages for loss of consortium, Flaminio sued Japanese Honda and American Honda, alleging that either the wobble was due to the defective design of the motorcycle, which should have been corrected, or the defendants should have warned users about the motorcycle's propensity to wobble. The jury exonerated Japanese Honda from all liability but found that American Honda had been negligent and that its negligence had been 30 percent responsible for the accident, and that Flaminio had also been negligent and that his negligence had been 70 percent responsible for the accident. Since under Wisconsin law, which the parties agree governs the substantive issues in this suit, a contributorily negligent plaintiff cannot recover any damages unless his negligence "was not greater than the negligence of the person against whom recovery is sought," Wis.Stat. Sec. 895.045; see, e.g., Brunette v. Employers Mutual Liability Ins. Co., 107 Wis. 2d 361, 364, 320 N.W.2d 43, 44 (Wis.App.1982), judgment was entered for the defendants, and the plaintiffs have appealed.

The substantive issues on appeal relate to Japanese Honda's liability for the allegedly defective design that allowed the motorcycle to wobble, or alternatively for failing to warn consumers of the danger of wobble. The issue with respect to the failure to warn is whether the judge should have given a strict liability instruction. He instructed the jury that if a product is dangerous when used as intended and the average consumer would not know this, "it becomes the manufacturer's or the distributor's duty to exercise ordinary care to give adequate warnings .... The failure to perform such a duty constitutes negligence." Flaminio argues that the judge should have instructed the jury instead that "A manufacturer or supplier of a product must give warning of any dangerous propensity of an article produced or sold by him inherent in the product or in its use of which he knows or should know, and which the user of the product would not ordinarily discover." This instruction, unlike the instruction the judge actually gave, did not refer explicitly to negligence or duty of care.

Since the jury found American Honda liable for the accident to Flaminio, it might seem irrelevant whether Japanese Honda should also have been found liable. But if, under different instructions, the jury had found Japanese Honda as well as American Honda liable, even if just for failing to warn of wobble, this might have affected the jury's apportionment of responsibility for the accident between Flaminio and the defendants. It is true that under Wisconsin's comparative-negligence statute (on which see the interesting recent discussion in Jensen v. Intermountain Health Care, Inc., 679 P.2d 903, No. 17980 (Utah S.Ct. Feb. 15, 1984)) Flaminio probably could get no damages unless he showed that he was less blameworthy than at least one of the defendants; it would not be enough that he was less blameworthy than both of them together. See, e.g., Wisconsin Natural Gas Co. v. Ford, Bacon & Davis Construction Corp., 96 Wis. 2d 314, 326-27, 291 N.W.2d 825, 831 (1980). We say "probably" because the fact that one of the defendants is the other's wholly owned subsidiary might conceivably make a difference. Compare Reber v. Hanson, 260 Wis. 632, 637-38, 51 N.W.2d 505, 507-08 (1952), with Reiter v. Dyken, 95 Wis. 2d 461, 467-68, 290 N.W.2d 510, 513-14 (1980). But we need not decide that question. It is enough that if the jury had found Japanese Honda liable, it might have concluded that Flaminio's fault was less than that of at least one of the defendants, and if so Flaminio would have gotten substantial damages.

The parties agree that Wisconsin law makes manufacturers and distributors strictly liable for personal injuries caused not only by defects in their products but also by inherent conditions that make the products unreasonably dangerous, though in the latter case the manufacturer's or distributor's duty is not to correct the condition but to warn potential users of the danger. But as we suggested recently in Gracyalny v. Westinghouse Elec. Corp., 723 F.2d 1311, 1317 n. 11 (7th Cir.1983), the strict liability duty to warn, as usually formulated (for example in DeLuryea v. Winthrop Laboratories, 697 F.2d 222, 228-29 (8th Cir.1983); Young v. Up-Right Scaffolds, Inc., 637 F.2d 810, 814 (D.C.Cir.1980); Werner v. Upjohn Co., 628 F.2d 848, 858 (4th Cir.1980), and Restatement (Second) of Torts Sec. 402A, comment j (1965)), is hard to distinguish in practice from the duty to warn imposed by a negligence standard. This is because the defendant, to be held strictly liable, must have been able to foresee that the product would be unreasonably dangerous unless there was a warning. "[T]he duty to warn [in a strict liability case] arises if the seller has, or should have, knowledge of a dangerous use." Krueger v. Tappan Co., 104 Wis. 2d 199, 206-07, 311 N.W.2d 219, 223 (Wis.App.1981).

The Wisconsin Supreme Court, in D.L. v. Huebner, 110 Wis. 2d 581, 614, 329 N.W.2d 890, 905 (1983), explicitly declined to say whether Krueger had been correctly decided; but we think that Krueger states Wisconsin law correctly at least on the point for which we have cited it. As is widely recognized in Wisconsin and elsewhere, "strict liability" is something of a misnomer in products cases. See Dippel v. Sciano, 37 Wis. 2d 443, 459-60, 155 N.W.2d 55, 63 (1967). There is liability only if a product is defective or unreasonably dangerous, and the concepts of "defect" and "unreasonableness" bring into play factors of cost and risk similar to those that determine negligence, an objective standard that is independent of what the particular defendant knew or could have done. "In defining unreasonably dangerous, a balancing test is mandated: if the likelihood and gravity of harm outweigh the benefits and utility of the product, the product is unreasonably dangerous." Brown v. Link Belt Division, 666 F.2d 110, 115 (5th Cir.1982); see also Epstein, Modern Products Liability Law 83 (1980). A similar balancing test is used in negligence cases. See, e.g., United States Fidelity & Guaranty Co. v. Jadranska Slobodna Plovidba, 683 F.2d 1022, 1025-26 (7th Cir.1982). As one court said recently, "in a defective design case, there is no practical difference between strict liability and negligence. The test for an 'unreasonably dangerous' condition is equivalent to a negligence standard of reasonableness ...." Birchfield v. International Harvester Co., 726 F.2d 1131, 1139 (6th Cir.1984). The leading treatise on tort law is in agreement: "The proof required of a plaintiff seeking to recover for injuries from an unsafe product is very largely the same, whether his cause of action rests upon negligence, warranty, or strict liability in tort." Prosser, Handbook of the Law of Torts 671 (4th ed. 1971); see also Howes v. Deere & Co., 71 Wis. 2d 268, 273-75, 238 N.W.2d 76, 79-80 (1976); Dippel v. Sciano, supra; Epstein, supra, at 40.

Of course there are some differences between negligence and strict liability in products cases. For example, when the defect is in a component and the manufacturer of the final product could not reasonably have been expected to discover the defect by inspecting or testing the component, the manufacturer would not be liable under a negligence regime for the defect, but under strict liability he would be. The parallel in a warning case would be where the element of danger that required a warning pertained to a component and the manufacturer could not reasonably be expected to discover the danger. But that could not be an issue with respect to Japanese Honda's liability in this case. If anyone should have known of the wobble danger it was Japanese Honda, not some component manufacturer. Yet it is only with respect to Japanese Honda's liability for failure to warn that the difference in the instructions could have affected the jury's verdict, since the jury found American Honda liable. Japanese Honda was liable for failing to warn of wobble if and only if its Gold Wing motorcycle was unreasonably dangerous without a warning, or stated otherwise if Japanese Honda should have known (whether it did or not) that a warning was necessary.

We need not go so far as to hold that the inquiry is identical in a case such as this whether the standard is denominated strict liability or negligence. But the difference if any is so small that the failure to give an explicit strict liability instruction is unlikely to be reversible error. Not surprisingly, there was little real difference between the instruction the district judge gave and the instruction Flaminio asked him to give. The judge instructed the jury that the defendants should be found liable if they had failed to exercise due care to warn consumers of any danger in the use of the product that the consumer might not be aware of. Although the instruction said nothing about what the defendants should have known regarding the danger and the consumer's awareness or unawareness of it, the reference to due care implied that they could be held liable only if they either knew or should have known that there was such a danger. Flaminio's proposed instruction required a warning "of any dangerous propensity" of the Gold Wing motorcycle which either of the defendants "knows or should know" of. This instruction, if given, would not have put before the jury a concept of strict liability distinct from the negligence instruction the judge did give; for without using the words "negligence" or "duty of care," the instruction would have made Japanese Honda liable only if the jury had found that Japanese Honda was negligent in failing to warn of the danger of wobble because it knew that there was such a danger and that users of the motorcycle would not realize and guard against it. Nevertheless, Flaminio's counsel was allowed in his closing argument to tell the jury that strict liability meant that even if the defendants had exercised all the care in the world they could still be held liable for selling an unreasonably dangerous product. And the defendants' counsel himself acknowledged to the jury that strict liability could be based "on either theory, the theory of defect and design [defective design?] and the theory of failure to warn."

None of this explains why the jury found the distributor liable for failing to warn, but not the manufacturer. Maybe the jury thought that it would have been redundant for both companies to issue warnings and that American Honda, being nearer the American consumer, was in a better position than Japanese Honda to know what warnings that consumer would need. But in any event there is no reason to think the jury would have reached a different conclusion about Japanese Honda's liability if Flaminio's instruction had been given. It too was phrased disjunctively, thus inviting the jury to distinguish between the manufacturer's and the distributor's duty to warn. And Flaminio does not argue that the jury verdict on liability (as distinct from damages, on which more later) is so internally inconsistent as to warrant a new trial. Compare Witt v. Norfe, Inc., 725 F.2d 1277, 1279-80 (11th Cir.1984).

The issue with respect to the allegation of defective design is whether the district court erred in excluding evidence (consisting of two blueprints) that, the plaintiffs say, shows that after the accident Honda, in an effort to reduce wobble, made the struts ("front forks") that connect the Gold Wing's handlebars to its front wheel two millimeters thicker. Rule 407 of the Federal Rules of Evidence makes evidence of subsequent remedial measures "not admissible to prove negligence or culpable conduct in connection with the event," but adds: "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 precautionary measures, if controverted, or impeachment." Flaminio argues that the blueprints were admissible under the exceptions for "proving ... feasibility of precautionary measures, if controverted," and for impeaching the defendants' evidence. But the first of these exceptions is inapplicable because the defendants did not deny the feasibility of precautionary measures against wobble. Their argument was that there is a tradeoff between wobble and "weave," and that in designing the model on which Flaminio was injured Japanese Honda had decided that weave was the greater danger because it occurs at high speeds and because the Gold Wing model--what motorcycle buffs call a "hog"--was designed for high speeds. The feasibility, as distinct from the net advantages, of reducing the danger of wobble was not in issue. As for the second exception, if the defendants had testified that they would never have thickened the struts on the Gold Wing the blueprints would have been impeaching. But the defendants offered no such testimony. Although any evidence of subsequent remedial measures might be thought to contradict and so in a sense impeach a defendant's testimony that he was using due care at the time of the accident, if this counted as "impeachment" the exception would swallow the rule.

Flaminio also argues that Rule 407 does not apply to strict liability cases. There is a conflict among circuits on the question. Most hold that it does. See Grenada Steel Industries, Inc. v. Alabama Oxygen Co., 695 F.2d 883, 886-88 (5th Cir.1983); Hall v. American Steamship Co., 688 F.2d 1062, 1066-67 (6th Cir.1982); Josephs v. Harris Corp., 677 F.2d 985, 990-91 (3d Cir.1982); Cann v. Ford Motor Co., 658 F.2d 54, 60 (2d Cir.1981); Werner v. Upjohn Co., supra, 628 F.2d at 856-57; Roy v. Star Chopper Co., 584 F.2d 1124, 1134 (1st Cir.1978). But the Tenth Circuit holds that it does not apply, see Herndon v. Seven Bar Flying Service, Inc., 716 F.2d 1322, 1331 (10th Cir.1983), at least where state law would, in a strict liability case, allow evidence of subsequent remedial measures, see Moe v. Avions Marcel Dassault-Breguet Aviation, 727 F.2d 917, 932 (10th Cir.1984). The Eighth Circuit's position is difficult to characterize. Compare DeLuryea v. Winthrop Laboratories, supra, 697 F.2d at 229, with Unterburger v. Snow Co., 630 F.2d 599, 603 (8th Cir.1980). This circuit was able to avoid the question in Oberst v. International Harvester Co., 640 F.2d 863 (7th Cir.1980), although Judge Swygert, in a separate opinion, argued that the rule is not applicable to strict liability cases, id. at 869-70. Even so, it might not be proper to allow evidence of subsequent remedial measures in a case submitted to the jury on both negligence and strict liability theories, as this case was; the prejudicial effect would be great even if the district court instructed the jury to consider the evidence only in relation to strict liability. But maybe this consideration would not be decisive. See Herndon v. Seven Bar Flying Service, Inc., supra, 716 F.2d at 1331. The exceptions in Rule 407 imply that the framers thought juries capable of limiting their consideration of evidence of subsequent remedial measures to issues other than the defendant's culpability.

In any event, we agree with the majority view that the rule does apply to strict liability cases. We are not persuaded by the purely semantic argument to the contrary that since "culpable conduct" is not the issue in such a case--the defendant is liable, at least prima facie, even if he is not blameworthy in the sense of being willful or negligent, provided that he caused the plaintiff's injury--the rule is inapplicable by its own terms. Wisconsin law rejects this argument in holding that a defendant's blameworthiness must, under the state's comparative-negligence statute, be compared with the plaintiff's blameworthiness in strict liability cases, even though the defendant was not blameworthy in a negligence sense. See Dippel v. Sciano, supra, 37 Wis.2d at 461-62, 155 N.W.2d at 64-65 (dictum). A major purpose of Rule 407 is to promote safety by removing the disincentive to make repairs (or take other safety measures) after an accident that would exist if the accident victim could use those measures as evidence of the defendant's liability. One might think it not only immoral but reckless for an injurer, having been alerted by the accident to the existence of danger, not to take steps to correct the danger. But accidents are low-probability events. The probability of another accident may be much smaller than the probability that the victim of the accident that has already occurred will sue the injurer and, if permitted, will make devastating use at trial of any measures that the injurer may have taken since the accident to reduce the danger.

The analysis is not fundamentally affected by whether the basis of liability is the defendant's negligence or his product's defectiveness or inherent dangerousness. In either case, if evidence of subsequent remedial measures is admissible to prove liability, the incentive to take such measures will be reduced. It is true that when liability is strict the defendant may have no incentive to take subsequent remedial measures even if Rule 407 would be applicable if he did. The accident may have been unavoidable at reasonable cost by taking greater care, and if so the defendant would have no incentive to take greater care. But he would still be liable for having caused the accident--that is what strict liability means. See Shavell, Strict Liability Versus Negligence, 9 J. Legal Stud. 1 (1980). The distinction is often stated in the following way: the focus of negligence is on the defendant's conduct, but the focus of strict liability is on the dangerousness of the product regardless of the defendant's conduct. That is why, when the standard is strict liability, a defendant may be held liable even though he was blameless, as in the case we put earlier where a component had a defect that the manufacturer of the final product--and defendant in the product liability suit--could not have discovered at reasonable cost. But this distinction does not justify a refusal to apply Rule 407 in product cases. In those cases where the defendant would have no incentive to take remedial measures anyway, because the accident was unavoidable, Rule 407 is academic; there will be, by assumption, no subsequent remedial measures. But in other cases, as should be apparent from our earlier remarks on the meaning of strict liability in products cases, the injurer would be held liable on a theory of strict liability even though the accident could have been avoided at reasonable cost by taking more care. Especially in a product case, the accident may have been readily avoidable either by eliminating some defect or by warning the consumer of some inherent danger, and in such a case the failure to apply Rule 407 might deter subsequent remedial measures just as much as in a negligence case. Cann v. Ford Motor Co., supra, 658 F.2d at 60. This is more than conjecture; the premise of Flaminio's offer of proof was that the accident was caused by a defect that could be and was later eliminated by a minor change in the design of the motorcycle. The policy of Rule 407 is applicable.

But Flaminio also argues that Rule 407 is inapplicable because this is a diversity case, and that Wisconsin's law on the admissibility of subsequent remedial measures should be applied instead. Although that rule is worded identically to Rule 407, see Wis.Stat. Sec. 904.07, the Wisconsin Supreme Court has held that it does not apply in the case of a mass-produced product. "Economic realities will set the course and these realities are that the sooner remedial measures are taken, the less costly the defect will be to the manufacturer." Chart v. General Motors Corp., 80 Wis. 2d 91, 102, 258 N.W.2d 680, 684 (1977). D.L. v. Huebner, supra, broadened this holding to all product liability cases in which at least one of the plaintiff's liability theories is strict liability. See 110 Wis. 2d at 610-14, 329 N.W.2d at 903-05. We find ourselves in respectful disagreement with the analysis in these cases. It is true that the benefits of subsequent remedial measures to the defendant in avoiding future accidents and their associated liability costs are greater, the larger the scale of the defendant's activity; but, by the same token, the costs of those measures to the defendant, in making it more likely that he will be forced to pay damages for accidents that occurred before the measures were adopted, are also greater. The effects of scale are symmetrical. In any event, Huebner dropped this rationale of Chart without proposing a substitute one. Huebner does express skepticism that defendants even know about the rule excluding evidence of subsequent remedial measures, see 110 Wis. 2d at 606, 329 N.W.2d at 902; but large manufacturers, such as Honda, which are defendants in hundreds or thousands of product liability suits, must know all about it.

But whether we agree or disagree with Chart and Huebner would be of no consequence at all if state rather than federal law (if inconsistent) determines the question of admissibility, as Flaminio argues with support from a considered dictum in a Tenth Circuit decision rendered after argument in the present case, Moe v. Avions Marcel Dassault-Breguet Aviation, supra, 727 F.2d at 932-33, though contrary to the implicit holdings of the many decisions such as Grenada Steel Industries, Inc. v. Alabama Oxygen Co., supra, that apply Rule 407 in diversity cases without considering what the state law on the admissibility of subsequent remedial measures might be. This is a difficult question. Having been enacted by Congress rather than promulgated by the Supreme Court pursuant to the Rules Enabling Act, 28 U.S.C. Sec. 2072, the Federal Rules of Evidence are not subject to the Act's proviso that rules promulgated under it "shall not abridge, enlarge or modify any substantive right ...." And we have held that Congress intended the Federal Rules of Evidence to apply in diversity as well as federal-question and admiralty cases. In re Air Crash Disaster Near Chicago, 701 F.2d 1189, 1193 (7th Cir.1983). But we must not forget that Erie R.R. v. Tompkins, 304 U.S. 64, 58 S. Ct. 817, 82 L. Ed. 1188 (1938), held that federal courts were constitutionally obligated to decide diversity cases in accordance with the rules of decision prescribed by state law (statutory or common), though the precise source of this obligation is not clear from the Supreme Court's opinion. See Wright, The Law of Federal Courts Sec. 56 (4th ed. 1983), for an excellent discussion. It is true that, as Erie itself suggests, the obligation is limited to matters of substantive law--"Congress has no power to declare substantive rules of common law applicable in a State," 304 U.S. at 78, 58 S.Ct. at 822. Otherwise the Federal Rules of Civil Procedure would be unconstitutional in diversity cases. But the Notes of the Advisory Committee to Proposed Rule 407 indicate that the rule is based primarily (in the Committee's words) "on a social policy of encouraging people to take, or at least not discouraging them from taking, steps in furtherance of added safety." See also McCormick, Handbook on the Law of Evidence 666 (Cleary 2d ed. 1972). As is apparent from Chart, Wisconsin has come to a different conclusion on this question, at least with respect to products liability cases; and it might appear that under Erie this judgment--a judgment on a matter of substantive policy rather than trial management--should be controlling in any case, such as this, where Wisconsin law rather than federal law supplies the standard against which the defendant's conduct is to be judged.

But this overlooks the fact that the substantive judgment that underlies Rule 407 is entwined with procedural considerations. It is only because juries are believed to overreact to evidence of subsequent remedial measures that the admissibility of such evidence could deter defendants from taking such measures. As the Advisory Committee and others have argued (see the discussion and references in 2 Wigmore, Evidence in Trials at Common Law Sec. 283 at pp. 174-75 (Chadbourn ed. 1979)), to infer negligence from such measures is to commit the fallacy, to which juries have long been thought prone, of believing that "because the world gets wiser as it gets older, therefore it was foolish before." Hart v. Lancashire & Yorkshire Ry., 21 L.T.R. (n.s.) 261, 263 (Ex.1869). Congress's judgment that juries are apt to give too much weight to such evidence is a procedural judgment (cf. In re Air Crash Disaster Near Chicago, supra, 701 F.2d at 1195), that is, a judgment concerning procedures designed to enhance accuracy or reduce expense in the adjudicative process. It is therefore well within the power of Congress to make for the federal courts with respect to any class of cases within their jurisdiction. Article III of the Constitution extends the federal judicial power to diversity cases, and Article I empowers Congress to make all necessary and proper laws for executing the specific powers granted to Congress, among which are the power to create lower federal courts to exercise the federal judicial power. If, therefore, as no one doubts, Congress may prescribe the qualifications of jurors in diversity cases (as it has done in the Jury Selection and Service Act of 1968, 28 U.S.C. Secs. 1861 et seq.), it may also decide what evidence should be excluded because its impact on those jurors is believed to be unduly prejudicial.

Although Rule 407 has substantive consequences by virtue of affecting incentives to take safety measures after an accident occurs, this just puts the rule in that borderland where procedure and substance are interwoven. The necessary-and-proper clause of Article I extends into the borderland, where are found "matters which, though falling within the uncertain area between substance and procedure, are rationally capable of classification as either." Hanna v. Plumer, 380 U.S. 460, 472, 85 S. Ct. 1136, 1144, 14 L. Ed. 2d 8 (1965). Although we agree with Judge Friendly that "it would subvert the scheme of the Constitution as this has come to be understood to read this auxiliary clause as a catchall empowering Congress to enact a code of private law applicable to all relations between citizens, subject to the sole qualification that the code applies only when a particular litigant can get into the federal courts under the changing rules that Congress lays down with respect to their jurisdiction," Friendly, In Praise of Erie--and of the New Federal Common Law, 39 N.Y.U.L.Rev. 383, 396 (1964), reprinted in Friendly, Benchmarks 155, 169 (1967), it would be melodramatic of us to view Rule 407 as one of the articles in such a code. It has substantive consequences, but that is true of many procedural and evidentiary rules. A notable example is Rule 35 of the Federal Rules of Civil Procedure, which authorizes compulsory physical and mental examinations of parties to federal diversity (as to other federal) cases. Rule 35 was held valid in Sibbach v. Wilson & Co., 312 U.S. 1, 61 S. Ct. 422, 85 L. Ed. 479 (1941), despite the fact that it both infringes on personal privacy and makes it harder for plaintiffs to get sizable damage awards in personal-injury cases. And the rule has only the authority of the Rules Enabling Act behind it. Despite the Act's explicit prohibition against abridging substantive rights, no rule promulgated under its authority has ever been invalidated, though some have been narrowed slightly. See Ragan v. Merchants Transfer & Warehouse Co., 337 U.S. 530, 533, 69 S. Ct. 1233, 1235, 93 L. Ed. 1520 (1949); Chesny v. Marek, 720 F.2d 474, 479-80 (7th Cir.1983), cert. granted, --- U.S. ---, 104 S. Ct. 2149, 79 L.Ed.2d ---- (1984). We are reluctant to cast a cloud over the whole federal rulemaking enterprise and open a new chapter in constitutional jurisprudence by holding that a procedural rule is beyond even the power of Congress to enact for application to diversity cases, because the rule affects substantive questions that the Erie doctrine reserves to the states.

But it can be argued that the procedural character of Rule 407 is compromised by the fact that not only a purpose but the primary purpose (insofar as one can judge from the Advisory Committee's Notes) was to promote safety--a purely substantive goal. But this is like arguing that the real purpose of the Act of July 31, 1912, 37 Stat. 240, which forbade importing "any film or other pictorial representation of any prize fight or encounter of pugilists, under whatever name, which is designed to be used or may be used for purposes of public exhibition," was to discourage prize fights. Of course it was; but this did not invalidate the statute. The Supreme Court rejected the proposition "that the motive of Congress in exerting its plenary power [to regulate interstate and foreign commerce] may be taken into view for the purpose of refusing to give effect to such power when exercised." Weber v. Freed, 239 U.S. 325, 330, 36 S. Ct. 131, 132, 60 L. Ed. 308 (1915); see also Hoke v. United States, 227 U.S. 308, 322, 33 S. Ct. 281, 284, 57 L. Ed. 523 (1913); Heart of Atlanta Motel, Inc. v. United States, 379 U.S. 241, 257, 85 S. Ct. 348, 357, 13 L. Ed. 2d 258 (1964). We must not push the analogy too far. The power to regulate procedure in the federal courts is quite dissimilar to the commerce power, so maybe a procedural regulation that was purely substantive in its purpose and impact could not be upheld under the former power. But Rule 407 is not based on substantive considerations only. An important though not the primary reason for the rule was distrust of juries' ability to draw correct inferences from evidence of subsequent remedial measures. Although it was a mild distrust, as shown by the exceptions built into the rule, it is enough to establish the rule's constitutionality in diversity cases.

The last two issues raised by this appeal can be disposed of quickly. Flaminio argues that because the jury assessed his wife's damages for loss of consortium above his own damages for pain and suffering ($500,000 versus $375,000) the verdict is inconsistent and he should get a new trial. Much more than a disagreement with the jury's judgment in assessing damages in regard to items of loss as devoid of objective guideposts as pain and suffering and loss of consortium would be necessary to upset a jury verdict as being internally inconsistent. See Canada Dry Corp. v. Nehi Beverage Co., 723 F.2d 512, 520 (7th Cir.1983). But in any event the record supports the assessment of higher damages for Mrs. Flaminio's intangible losses than for her husband's. Although a paraplegic, he has adjusted well to his condition and is able to get around to the extent that he even goes hunting. His wife on the other hand had to quit her job and become his full-time personal attendant. A rational trier of fact could regard her lot as worse than his, as apparently this jury did.

Finally, Flaminio argues that the trial judge unduly curtailed his opportunity to present his case by announcing at the outset of the trial that the trial would be allowed to take only 33 hours--18 for the plaintiffs and 15 for the defendants. Although in this era of crowded district court dockets federal district judges not only may but must exercise strict control over the length of trials, and are therefore entirely within their rights in setting reasonable deadlines in advance and holding the parties to them, see, e.g., MCI Communications Corp. v. American Tel. & Tel. Co., 708 F.2d 1081, 1170-72 (7th Cir.1983), we disapprove of the practice of placing rigid hour limits on a trial. The effect is to engender an unhealthy preoccupation with the clock, evidenced in this case by the extended discussion between counsel and the district judge at the outset of the trial over the precise method of time-keeping--a method that made the computation of time almost as complicated as in a professional football game. But our disagreement with the district judge's method of economizing on trial time does not warrant reversal of his judgment. The 18 hours that the plaintiffs were given to put in their case were not an unreasonable period in relation to the complexity of the issues, and in any event the plaintiffs have failed to indicate what evidence they would have put in, or cross-examination they would have conducted, if they had had more time. The main thing they were not allowed to do was to explore the defendants' alleged remedial measures, but they were prevented from doing that for reasons other than time. We trust, however, that in the future the able district judge will not try to slice the loaf so thinly.

AFFIRMED.