Concept
Why the Rule Exists
Remedial measures
Why the Rule Exists
Rule 407 lists what a court must exclude and what it may still admit, but it never says why.1 The reason matters because it governs the cases the text leaves open.
The Advisory Committee gave two reasons, and they differ in kind.2 The first concerns the evidence. A measure taken after an injury is not an admission of fault, because the conduct is equally consistent with an accident no one could have prevented or with the plaintiff’s own carelessness. The second concerns conduct outside the courtroom. The law wants people to make hazards safer, or at least not to discourage them, and a party will hesitate to fix a hazard if the fix can later go to a jury as proof of fault.
The Committee ranked the two. It called the safety-policy ground the “more impressive” one and conceded that the first ground alone would not support exclusion under a liberal theory of relevancy, “as the inference is still a possible one.”3 Evidence of modest weight ordinarily goes to the jury, which decides how much it proves. A categorical rule needs more, and the policy of encouraging safety supplies it. The evidentiary reason still does work: it shows how little probative value the rule gives up.
These reasons give you a test for unfamiliar facts. Ask whether the evidence would let a jury read a repair as a confession it does not amount to, and whether admitting it would give someone a reason to delay a safety measure. The answers tell you how far the exclusion should run and how far each exception should reach. Courts divide, for example, over how broadly to read the feasibility exception, and the broader reading circumscribes the exclusion.4 The evidentiary reason comes first below, and the incentive reason second.
Ambiguous Evidence of Fault
The first reason attacks the inference a plaintiff needs. To use a repair as proof of fault, the jury must move from “the owner fixed it afterward” to “the earlier condition was unreasonably dangerous,” and that step holds only if people repair things mainly because they were dangerous before. The Advisory Committee rejected the premise: later conduct is “equally consistent with injury by mere accident or through contributory negligence,” so it is not in fact an admission.5 A repair may also reflect nothing more than a desire to exercise the highest care, which is not an admission of negligence or culpable conduct.6
Baron Bramwell stated the point in the form the Committee quoted. The rule, he said, rejects the notion that “because the world gets wiser as it gets older, therefore it was foolish before.”7 An accident supplies knowledge that did not exist beforehand, and a person who acts on it has not conceded that the earlier conduct fell short.
A plaintiff cannot avoid this reasoning by relabeling the repair. In one case the plaintiff offered evidence of a modified jack pin as an admission by a party-opponent. The court held that non-hearsay is not automatically admissible, that it remains subject to other rules, including the state’s version of Rule 407, and that the plaintiff could not escape the exclusion by calling the evidence an admission.8
Jurors compound the problem because they may give a repair more weight than its ambiguity deserves. The rule’s drafters treated corrective steps as “likely to be overvalued by the jury.”9 In Duchess v. Langston Corp., the court observed that jurors can treat a remedial measure as an effective concession that the product, in its earlier form, was defective, and it found the danger illustrated by the plaintiffs’ own arguments, which they presumably would make to a jury.10 Evidence of this kind invites the jury to try what the defendant did afterward instead of what the condition was when the injury occurred.
The ambiguity reason does not carry the rule alone. The Advisory Committee conceded that, under a liberal theory of relevancy, this ground would not support exclusion because the inference remains possible.11 The Duchess court accepted that remedial measures have some relevance, though it may be limited, and it acknowledged that, apart from social-policy concerns, leaving the matter to Rule 403 weighing in each case might be the preferable course. It held that the policy considerations tip the balance toward a general rule of exclusion.12
The ambiguity also helps explain why the exceptions exist. The rule bars the evidence when it is offered to prove negligence, culpable conduct, a defect, or a need for a warning, and it permits another purpose, such as impeachment or, if disputed, proving ownership, control, or the feasibility of precautionary measures.13 A defendant who denies owning a stairway and then replaces its broken rail has produced evidence of ownership whose force does not depend on reading the repair as a confession of earlier danger, because people rarely pay to repair what belongs to a stranger. The jury concern does not disappear, though. Evidence offered for a permitted purpose remains subject to Rule 403, and if jurors’ tendency to take the repair as an admission of fault substantially outweighs its legitimate probative value on ownership or control, the court may exclude it.14 The later discussion of Rule 403 balancing develops that point.
Encouraging Safety Measures
The second reason looks past the courtroom to what a party does about a hazard once someone has been hurt by it. One of tort law’s primary goals is to encourage people to take steps to avoid accidents, and one way to serve that goal is to adopt rules that do not penalize such steps.16 Evidence that a defendant improved a condition after an accident is relevant, because it tends to show the defendant thought the condition dangerous when the accident happened. But a rule that let the injured party prove the improvement might discourage the defendant from making it and so raise the risk of future accidents. Avoiding that discouragement is the main purpose of Rule 407, which serves it by excluding the evidence when it is offered to prove fault.17 The Advisory Committee described the principle as encouraging people to take, or at least not discouraging them from taking, steps toward added safety, and it noted that courts had applied it to subsequent repairs, installed safety devices, changed company rules, and discharged employees.18
The mechanism runs through a choice. After an accident, a party can fix the hazard now or leave it until the case ends. If the fix can go before the jury as proof of fault, waiting costs nothing in the lawsuit while fixing costs something, so a party thinking only about the lawsuit has a reason to wait while the hazard stays in place for the next person.19 The court in Flaminio v. Honda Motor Co. supplied the arithmetic. One might think it immoral, even reckless, for an injurer alerted to danger by an accident to leave it uncorrected. But accidents are low-probability events, and the chance of a second one may be far smaller than the chance that the first victim will sue and make devastating use at trial of any measure taken since.20
That effect depends on a belief about jurors. As one court put it, only because juries are believed to overreact to evidence of subsequent remedial measures could the admissibility of such evidence deter defendants from taking them.21 The two reasons for the rule therefore fit together: the jury’s tendency to overvalue a repair, which makes the evidence unreliable, is also what creates the incentive to delay it.
A narrower argument sits beside the incentive argument. Using a repair against the person who made it amounts to punishing the actor for laudable conduct, and the rule also rests on the dubious probative value of such evidence as a sign of unreasonable danger when the earlier accident occurred.22 The two arguments reinforce each other: because a repair proves so little about the earlier condition, shielding it costs the litigation system little.
Whether the incentive is real for a large manufacturer is contested. Take an automobile maker that learns a design element has led to accidents. One can predict that, to prevent more accidents, it will change the design even if it knows that evidence of the change could be used in suits over earlier accidents, because a simple cost-benefit analysis points that way.23 If so, the exclusion adds little to its motive to act. Ault v. International Harvester Co. reasoned from the scale of a mass producer’s operations, and Flaminio disagreed: exposure to liability from future accidents if a defect stays unfixed is offset by exposure to liability from past accidents if the defect is fixed and proof of the fix is admitted, so the “effects of scale are symmetrical.”24 The plaintiff’s own offer of proof in Flaminio assumed that the accident came from a defect later eliminated by a minor design change, and the court found the policy of Rule 407 applicable.25 Another court named the encouragement of safer products as the rule’s second goal, reasoning that manufacturers would hesitate to keep updating safety features if each update could be used to show that earlier models were defective.26
Courts apply the rule even though the premise is contested. Most states apply the exclusion across the board whenever a party takes a remedial measure and the evidence is offered for a forbidden purpose, and the federal rule was amended to make clear that it covers product liability cases.27 Ault stands among the minority that decline to do so, a split the later discussion of product cases takes up.
The incentive premise also frames a question taken up under “Who Took the Measure.” The rule protects a party whose own decision to repair would otherwise be chilled, and a repair by a nonparty cannot be used to prove that nonparty’s liability in this suit. Diehl v. Blaw-Knox presents the question: the post-injury change came from someone who was not a party.