When an alleged hazard exists and causes a person to be injured, then maybe a business might think about fixing it. But would the fix amount to a confession and cause the business to own that prior liability? That’s the question that motivates Rule 407 in the Federal Rules of Evidence, and its many state-law versions: To prevent jurors from applying hindsight and giving too much weight to remedial measures as proof of liability, plaintiffs are generally prohibited from introducing that evidence. And defendants generally like that rule because they’re wary of providing jurors with any shortcut to liability. But it may not be so simple. Evidence of subsequent measures in some situations may actually help the defendant.

The two-edged nature of subsequent remedial measures is the subject of a research article from Bernard Chao, a University of Denver Law Professor, and law student Kylie Santos (Chao & Santos, 2019) soon to be published in the Missouri Law Review. Conducting a series of experiments, they used rich fact patterns, including video-recorded arguments from both sides, as well as jury instructions, based on two case scenarios. One fact pattern involved a suit over an individual’s fall down some stairs, while a second focused on a snowboarder’s defective design suit against a snowboard manufacturer. Varying the inclusion of evidence on subsequent remedial measures, as well as instructions on the same, the researchers presented the scenarios to a large pool of 1,700 mock jurors who rendered verdicts on liability, contributory negligence, and damages. The bottom line: Knowing about later fixes helps the plaintiff on liability, but may sometimes help the defendant on damages. Because there are some exceptions to Rule 407, it can be a strategic call for both plaintiffs and defendants. I’ll take a closer look at these results in this post.

Here’s When It Hurts Defendants

The most expected finding of the study is the confirmation that evidence on remedial measures helps to drive liability verdicts in the plaintiff’s direction. As the research article concludes, “Our results confirm the widely held notion that evidence of subsequent remedial measures leads jurors to find defendants liable at a higher rate.” When mock jurors heard the evidence of fixes the defendants had applied at a later time, the plaintiffs’ win rates went up. This stands to reason based on the tendency for jurors to think that, in effect, if it was fixed, then it must have been broken. In addition, it also serves to undercut the credibility of defense arguments denying liability, with the jurors thinking that if the defendant actually believed it wasn’t a danger, then they would not have made any changes.

Here’s How to Instruct Against that Harm

The study also found that jury instructions can reduce, but not eliminate, the bias on liability. In order to assess the role of the rules of evidence, the researchers compared instructions that merely instruct to instructions that also give a reason.

The simple ‘limited purpose’ instruction was as follows:

“The evidence was only admitted for the limited purpose of challenging the defendant’s argument that this defendant’s original design was the safest one possible. You must consider it for that purpose only and may not consider it as evidence that the product was defective.”

While a second version added an explanation:

“We have this rule for two reasons. First, we have this rule because we want to encourage people to fix problems and the law will not use evidence of such fixes against them later in court. Second, we are concerned that the jury may overweigh evidence of remedial measures. Although it may tend to show that the original version was defective, it is possible that companies can improve products that were not defective.”

Ultimately, I am not sure I’d advise a defendant to agree with that latter language. The phrase “may tend to show that the original version was defective” seems to give away too much. It would be more neutral in my view to say, “Evidence on remedial measures demonstrates an ability to change or improve, but does not bear one way or the other on whether the original design was defective.”

However, in one experiment, including the instruction with an explanation significantly reduced, but still did not eliminate the pro-liability bias of the subsequent remedial measures.

And Here’s When It Helps Defendants

This was apparently the first study to test the effects of subsequent remedial measures on damages. To some extent, at least, the information that the problem had been addressed seemed to reduce the motivation to award high damages. In the snowboard fact pattern, when mock jurors learned that the design had been changed, damages awarded against the defendant were reduced.

Given that the damages in this case were compensatory rather than punitive, that might not be strictly proper. “After all,” the authors note, “evidence of what the defendant did after an accident should not affect how juries view the gravity of the plaintiff’s injury or what damages should be awarded to compensate for that injury.” However, there is a well-known tendency for jurors to vary damages based on the perceived ‘blameworthiness’ of the defendant, and to treat damages more generally as an act of ‘sending a message’ and not simply compensating for an injury. In this case, learning that the manufacturer had changed the design encouraged mock jurors to see the defendant as less blameworthy and less in need of a message. For that reason, the information on subsequent remedial measures reduced damages, counteracting the increased liability finding.

And Here’s Why You Need to Test Your Own Case

The study in this case tested two scenarios with different results. The size of each of the effects varied, and most importantly, sharing the subsequent measures helped the defendant in one scenario, but not in the other. The researchers attribute this difference to the fact patterns themselves, which is why they used two rather than one.

“Cases in our legal system are complex and decision making often depends on the particular facts being tested,” they note. “To the extent any study suggests particular insights, the results need to be tested in different contexts to determine how well the insight holds up when combined with different factors.”

The implication is that, while research of the kinds shared in this blog can be helpful in forming general expectations, there is no substitute for testing what the reactions will be to your own case. That is a great reason to consider a community survey, a focus group, or a mock trial.