How Evidence of Subsequent Remedial Measures Matters.

• Author: Chao, Bernard

TABLE OF CONTENTS ABSTRACT 609 INTRODUCTION 613 I. BACKGROUND 615 A. Federal Rule of Evidence 407 615 B. Prior Studies 621 C. Damages 624 D. Hypothesis 625 II. EXPERIMENT 1: PREMISES LIABILITY 626 A. Scenario 1: The Basic Case 627 B. Scenarios 2 and 3: Not Feasible & Remedial Measures 627 C. Scenarios 4 & 5: Limiting Jury Instructions 628 D. Respondents 629 E. Results 630 1. Liability (Staircase) 630 Table 1 631 Staircase Experiment, Liability 631 2. Damages (Staircase) 632 Table 2 633 staircase Experiment: 633 Damages and Contributory Negligence 633 3. Case Expected Value (Staircase) 634 4. Summary (Staircase) 634 III. EXPERIMENT 2: PRODUCTS LIABILITY 635 A. Scenario 1: The Basic Case 636 B. Scenario 2: The Presentation of Subsequent Remedial 637 Measures C. Scenarios 3 and 4: Limiting Jury Instructions 638 D. Respondents 639 E. Results 639 1. Liability (Snowboard) 639 Table 3 640 Snowboard Experiment, Liability 640 2. Raw Damages (Snowboard) 641 Table 4 642 Snowboard Experiment: 642 Damages and Contributory Negligence ("CN") 642 3. Contributory Negligence (Snowboard) 643 4. Net Damages (Snowboard) 643 5. Case Expected Value (Snowboard) 644 6. Findings 644 IV. DISCUSSION 645 A. Implications 645 1. Policymakers 645 2. Litigants 646 3. Researchers 647 B. Limitations 648 CONCLUSION 650 APPENDIX A 651 Table A1 651 Table A2 651 APPENDIX B 654 Snowboard Analyses 654 Table B1 654 Table B2 655 Table B3 656 Table B4 657 Table B5 658 APPENDIX C 659 INTRODUCTION

Federal Rule of Evidence 407 ("Rule 407") prohibits plaintiffs from introducing evidence of subsequent remedial measures to show that the defendant is to blame. (1) Among its purported justifications, the rule prevents hindsight bias from unduly influencing jury decisions. (2) Nonetheless, plaintiffs often take advantage of the rule's numerous exceptions to introduce evidence of remedial measures for other purposes (e.g. to prove feasibility). Fearing that the exceptions could swallow the rule, some courts will even exclude evidence that fits into one of these exceptions because it is ostensibly too prejudicial. (3) Alternatively, other courts instruct juries that they should only use the evidence for the limited permissible purpose, but not for proving blameworthiness. (4)

This complex scheme makes several assumptions about how evidence of remedial measures and the accompanying limiting instructions influence juries. Although many studies have examined hindsight bias in other contexts, and one older study looked at Rule 407 in particular, these studies typically used short, written vignettes with small sample sizes. Moreover, none of these studies examined how subsequent measures impact damages. We sought to test these concepts in robust fashion by conducting two separate experiments using videos that included rich fact patterns with arguments from both parties and jury instructions. In the end, over one thousand seven hundred mock jurors rendered verdicts on liability, contributory negligence, and damages.

As expected, evidence of subsequent remedial measures helped plaintiffs win more often. (5) But surprisingly, our results also suggested that taking remedial measures may lower damages, thereby counteracting the increased liability findings. (6) We also studied the efficacy of two limiting jury instructions. In one experiment, a limiting instruction with an explanation reduced but did not eliminate the effects of evidence of subsequent remedial measures. The instruction with explanation was also consistently more effective than the simple limiting instructions, but these results were not statistically significant. We hypothesize about the potential reasons for our various results and discuss what they mean for policymakers, litigants, and future researchers that may wish to explore this subject in more depth.

This Article proceeds in four parts. Part I explains Rule 407 and its underlying justifications. In brief, the rule prohibits the introduction of evidence of subsequent remedial measures to show culpability. The thinking behind the rule is that hindsight bias causes juries to give too much weight to evidence of subsequent remedial measures, a result that would be unfair to defendants. However, the rule is riddled with exceptions that allow the evidence to be used for other purposes. But even when an exception applies, some courts will not admit the evidence of subsequent remedial measures because it is overly "prejudicial." Other courts issue limiting instructions to the jury telling them that they can use the evidence for one purpose but not another. This complex scheme makes several assumptions about the effect of evidence of subsequent remedial measures and limiting instructions. part i then goes on to explain that few studies have examined how Rule 407 affects decision-making. To the extent that studies do exist, they have only considered liability decisions. No studies have looked at how evidence of remedial measures might affect decisions on damages.

Part II then explains our first experiment (the staircase experiment). in this negligence case, the plaintiff hurt herself falling down overly steep stairs. We created five versions of this case, cumulatively layering on a new manipulation for each subsequent version: basic case (version 1); defendant argued that it was not feasible to make the staircase safer (version 2); plaintiff introduced evidence that the defendant later added landing (version 3); judge provided a simple limiting instruction (version 4); and judge explained the reason for the limiting instruction (version 5). For each of these versions, we asked mock jurors to determine liability and if applicable, contributory negligence and damages.

Part II then describes the results of the staircase experiment. As expected, evidence of subsequent remedial measures increased plaintiff's win rate. The limiting jury instructions with explanation reduced liability findings, while a simple instruction trended in the same direction but was not statistically significant. We did not observe any significant damages effects, but the limiting instruction with explanation appeared to be trending lower. We only mention the null damages findings because of the more significant results found in our second experiment.

Part III then explains our second experiment (the snowboard experiment). In this products liability case, the plaintiff was hurt during a snowboarding accident. He sued the snowboard manufacturer for a defective design. While we manipulated this case in two dimensions, four basic versions of this case are relevant to this article. (7) Again, we cumulatively layered on a new manipulation for each subsequent version: the defendant argued that it made the design as safe as it could while retaining the ability for customers to choose different bindings (version 1); plaintiff introduced evidence that the defendant later made a safer design (version 2); judge provided a simple limiting instruction (version 3); and judge explained the reason for the limiting instruction (version 4). Again, mock jurors were asked to determine liability and if applicable, contributory negligence and damages.

Part III then describes the results of the snowboard experiment. Like the staircase experiment, the snowboard experiment confirmed that evidence of subsequent remedial measures increased plaintiffs' win rate. Although the impact was smaller than we saw in experiment 1, it was still sizable and statistically significant. Interestingly, when the mock jurors saw evidence of subsequent remedial measures, the plaintiff's win rate increased, but net damages (raw damages discounted for findings of contributory negligence) decreased. Although the causal mechanism is unclear, we hypothesize that jurors thought the defendant's decision to take remedial measures made the defendant less blameworthy and therefore deserving of smaller damages. Because liability and damages went in opposite directions, we saw no statistically significant differences between any of our scenarios when we examined the case expected value. This effect was only present in the snowboard case. This suggests that, at least in some cases, evidence of subsequent measures may not harm defendants as much as previously thought.

Additionally, while the limiting instruction with explanation appeared to lower liability findings, that result was not statistically significant in the snowboard experiment. Moreover, the simple limiting instruction appeared to have no effect whatsoever. Finally, in Part IV, we describe implications of these different findings and the limitations of our study.

1. BACKGROUND

   1. Federal Rule of Evidence 407

      Rule 407 of the Federal Rules of Evidence prohibits the introduction of subsequent remedial measures to prove blameworthiness in negligence and products liability cases, but the rule allows such evidence to be admitted for other reasons, including to prove the feasibility of such remedial measures. (8) Most states have identical or substantially similar rules to the federal rule. (9) Although the rule was originally limited to cases involving claims for negligence, many state courts began applying this rule to products liability cases, which is a strict liability tort. (10) In 1997, Federal Rule of Evidence 407 followed suit, adding language that also prohibited the use of evidence of subsequent remedial measures to prove "a defect in a product or its design; or a need for a warning or instruction." (11) Many, but not all, states have expanded the rule to products liability. (12)

      There are two articulated justifications underlying Rule 407. First, the rule encourages accused defendants to take steps to improve safety without fear that these steps will be used against them in court. (13) Second, and perhaps more controversially, some commentators have questioned the probative value of evidence of subsequent remedial measures. (14) While there is widespread agreement that...