Don’t miss the consultant responses below: Ken Broda-BahmAlison K. Bennett.


1. Introduction

Many hours away from the nearest land mass, a cruise ship turns over due to the rough sea. Stan and six other passengers make it onto a small lifeboat. However, the lifeboat is over its maximum capacity and it is quickly filling up with water. Soon the lifeboat will sink altogether. So Stan pushes an elderly person off the lifeboat and into the rough waters. The elderly person drowns, while the lifeboat manages to stay afloat and the other five passengers survive.

However hypothetical this situation may seem, it embodies a more fundamental issue that arises commonly in the law: whether, and when, an act of harm can be justified. It is a plain fact that Stan killed a person—the difficult and pivotal issue to assess is whether he acted with reasonable justification. This same question has been studied intensively over the last fifteen years by psychologists: How do humans go about deciding when harmful behavior is justified by the benefits it brings about? Of course, the aim of the psychological research is not to prescribe how people should answer that question; rather, it is to describe the mental machinery that they use, whether rightly or wrongly.

A remarkable body of evidence indicates that our minds are often divided between two types of machinery, reason and emotion, which together shape the way we think and behave. The influence of reasoning processes on our moral thinking can often be quite simple: All else being equal, we tend to favor whatever course of action leads to the greatest amount of benefit in the long run (Bartels, 2008; Greene, Nystrom, Engell, Darley & Cohen, 2004; Moore, Clark & Kane, 2008; Paxton, Ungar & Greene, 2012). So when we reason about cases like the lifeboat dilemma, we tend to judge that Stan did the right thing in saving more lives. But the influence of emotions on moral judgment is far more fickle. For instance, we tend to have a stronger emotional reaction to harming a person in an up-close and personal way (for instance, pushing them off a boat) than in a distant and impersonal way (for instance, failing to keep sufficient life-preservers on board a boat; Crockett, Clark, Hauser & Robbins, 2010; Greene, Sommerville, Nystrom, Darley & Cohen, 2001; Koenigs et al., 2007; Youssef et al., 2012). So when emotions are guiding our thinking we are more likely to condemn what Stan did, even though it was the lesser of two evils.

Understanding the contours of our moral emotions is a matter of manifest practical significance, influencing the way we make decisions as citizens but also as policy-makers, judges, advocates and jurors. In our research, we distinguish two hypotheses about the basis of our emotional aversion to harm. On the one hand, moral condemnation is often thought to derive from empathy toward the innocent victim. This is a popular assumption in moral philosophy, and is present also in the Golden Rule (i.e., “Do unto others as you would have them do unto you”). According to this perspective, one focuses on the victim’s pain and suffering and subsequent feelings of empathy drive the conclusion that what Stan did to her was wrong. In our research, we draw attention to an alternative possibility: When one considers an immoral act, perhaps one actually focuses on the aggressor’s deed: i.e., one imagines what it would be like to forcefully push an elderly lady to her death. This process results in a feeling of aversion to the action itself, which motivates the attitude of moral condemnation (see Miller & Cushman, 2013). In sum, when we condemn a violent act carried out for the greater good, are we moved mainly by the victim’s suffering or, rather, by an aversion to what the aggressor did?

2. Findings

Our first challenge was to devise a method to separate harmful actions from their outcomes in order to isolate the corresponding aversions they are proposed to elicit. Our purpose in doing so was to differentiate individuals with high levels of action aversion from those with low levels of action aversion, and likewise for high versus low levels of outcome aversion, and then compare their moral judgments. So, we compiled a list of hypothetical scenarios in which apparently harmful actions are performed in special circumstances where they do not lead to harm: such as ‘stabbing a fellow actor in the neck during a play using a stage knife with a retractable blade,’ or ‘shooting a bullet at a consenting friend while he’s behind a bulletproof glass’. Conversely, we also compiled a list of hypothetical scenarios in which an individual is harmed but this harm is not caused by someone else’s action: for example, ‘seeing someone step on broken shards of glass,’ or ‘hearing a frightened child crying’. We asked participants to tell us how “upsetting” they would find each of these situations, and thus we were able to separately measure people’s sensitivity to harmful actions versus outcomes.

Participants were also instructed to make a number of judgments about the morality of killing one person to save many, in situations like the lifeboat case presented above. We then asked: Are participants’ judgments about these cases predicted by their aversion to actions or to the outcomes they cause? In a first experiment, we found that both measures predicted the extent to which participants condemned the moral dilemmas. Moreover, analyses showed that both measures made independent and complementary contributions to people’s moral judgments. This suggests that when judging the morality of a violent act one is moved not only by a concern for the harm being done, but also by an aversion to the action being performed.

Our first set of participants filled out both sections of the survey in a single testing session. This led to the worry that participants might have grasped the aim of our study and subsequently altered their responses. So, in our next experiment, we recruited a group of participants who had judged a set of moral dilemmas approximately two to three years ago. We asked these participants to complete our measures of action and outcome aversion. This time we found that, whereas action aversion remained strongly correlated with moral judgment, participants’ aversion to outcomes did not. Together with the results of our first experiment, these findings suggest that the aversion to harmful action plays a remarkably robust role, and perhaps even a greater role than empathy, in the moral condemnation of violence.

Even though our action aversion items described behaviors that were indeed physically harmless, it occurred to us that participants’ aversions may not have been due strictly to the performance of the action. ‘Making rude gestures at a friend behind his back’ does not bring about harm to the friend. However, it may result in other bad outcomes, like the disapproval of potential onlookers, which participants may have been concerned about. It is possible, therefore, that our ‘action’ aversion items were in fact eliciting certain kinds of outcome-based concerns. So, in our third experiment, we took extra care to select three items which we thought precluded these concerns altogether.

In order to obscure the purpose of our research, we made two further amendments to the design of our study. First of all, rather than having participants complete both sets of items, we had one group complete our action aversion items and another, separate group complete the outcome aversion items. Second, we placed these three items of interest (“How upset would it make you to see someone shut their own finger in a car door?” or “How upset would it make you to curse angrily at an old woman as part of a movie script?”) within a longer list including 21 “distractor” items, like “How upset would it make you to get a flat tire on the way to your wedding?” or “How excited would you be if scientists discovered life on another planet?”. This time the results strongly favored a selective role for action aversion. In other words, when participants faced situations where someone actively harmed another in order to save a greater number of lives, their moral judgments were predicted by how they would feel performing harmless acts with violent features, but not by how upsetting they would find it to witness harmful outcomes.

Across the previous experiments we observed a robust relationship between people’s aversion to harmful action and their tendency to condemn harm. Yet, in our research, we were interested in making a further causal inference: i.e., we wanted to know whether the aversion to action influences ourmoral judgments about it. Our final experiment was designed to test this related causal claim. To do so, we asked people to judge the aversiveness and moral permissibility of different methods of mercy killing. Our thinking was the following: Mercy killings vary in the amount of suffering the patient experiences prior to his death: a person infected with anthrax will die after very great and prolonged suffering, while electrocution may yield minimal suffering and an almost sudden death. They vary also in the degree to which the mercy killing would be disturbing to perform, even in a pretend context. Pretending to slit another person’s neck would be much more disturbing than simply administering a fake poisonous solution, even if one knew that both actions would cause no harm. So, we asked a group of participants how upsetting these mercy killings would be to perform (as part of a movie plot), and a separate group of participants to rate how much suffering the patient would experience. Averaging these responses gave us an approximation of the level of action and outcome aversion for each method of mercy killing. A third group of participants judged how morally permissible it would be to conduct each of these kinds of mercy killings.

Our results showed that both the aversive character of the performance of a mercy killing and the degree of suffering experienced by the patient strongly predicted the moral wrongness of the mercy killing. In fact, these effects appeared to be independent of each other and explained almost all the variability in people’s moral judgments. Additionally, unlike previous experiments, this study enabled us to directly link the action and outcome aversion elicited by an action to moral judgments about that same action. This suggests that the degree of suffering attributed to the victim, along with the aversion associated to the performance of a harmful action, shape our moral judgments about it.

3. Implications

These results cast some doubt on prevailing assumptions about morality. It is widely believed that the opposition to harmful behavior is motivated by an empathic feeling towards the victim. Here we uncover a different, and perhaps even greater, influence on our moral thinking, based in an aversion to the aggressor’s action. In conjunction with other recent studies, our results suggest that when judging third-party infractions, we imagine what it would feel like to perform the behavior oneself. In some cases, this process yields a feeling of aversion that contributes to our judgment that the behavior was wrong: “If it feels bad to me to do it, it’s wrong for you to do it”.

This may have some important implications for litigation in criminal law. In addition to the harm done to the plaintiff, jurors frequently take into account aspects of the manner in which the harm was done. Some of these features of a crime may reasonably aggravate the offense, by reflecting the defendant’s mens rea, ill intentions or criminal character. But, as we saw, a juror may also condemn a criminal offense more severely as a result of a personal feeling that it would be disturbing to do what the defendant did. This is remarkable in at least three respects.

First, certain kinds of cases that are psychologically disturbing may be criminalized more severely than other cases that are not equally disturbing, even when the latter bring about more harm. Second, sentencing for aggressive crimes may depend to a surprising degree on how much the juror is upset by imagining the defendant’s crime. Indeed, ongoing research finds reliable differences in the degree to which people use their own feelings of aversion as a basis for judging others (Hannikainen, Miller & Cushman, in prep.) Therefore, more ‘emotional’ jurors may condemn these psychologically disturbing, violent cases more severely than judges who are, at least in theory, less easily swayed by their emotions.

Third, our findings suggest that jurors’ decisions may be readily influenced by an attorney’s attempt to focus attention on the performance of the harmful action. It is common in courtroom settings for the prosecutor to direct the juror’s attention towards the victim’s damages (or for the defense attorney to direct juror attention to the plaintiff’s damages). This is done, of course, in order to elicit his or her feelings of sympathy, with the goal of winning the case, aggravating the sentence, or increasing the settlement. Our research indicates that directing attention to the defendant’s wrongdoing and, in particular, to the feelings associated with performing his or her action, may influence the juror at least as much. This influence on decision-making is relatively less recognized among psychologists, and may perhaps be more overlooked in the courtroom as well.

Ivar Hannikainen [] recently graduated with a PhD in Philosophy from the University of Sheffield, UK. His academic interests span philosophy and psychology, with a focus on issues relating to morality and decision-making. Ivar’s dissertation examines the influence of emotional processes and of rational thought on moral decision-making and political orientation. You can review his research and contact information on his webpage at

Ryan Miller [] is currently a PhD student in the Department of Cognitive, Linguistic, and Psychological Sciences at Brown University. His research focuses on understanding how emotions like harm aversion and empathy influence moral judgment, as well as identifying moderators of their influence (e.g. presentational format, cognitive reflection, etc.). Ryan is the recipient of a National Science Foundation Graduate Research Fellowship, and he received his BA in Psychology from Yale University. You can review Ryan’s research and recent publications at

Fiery Cushman is Assistant Professor of Psychology at Harvard University, where he directs the Moral Psychology Research Laboratory. His research investigates the cognitive mechanisms responsible for human moral judgment, along with their development, evolutionary history and neural basis. He received his BA and PhD from Harvard University, where he also completed a post-doctoral fellowship. He served as Assistant Professor of Cognitive, Linguistic and Psychological Sciences at Brown University from 2011 to 2014.


Bartels, D. M. (2008). Principled moral sentiment and the flexibility of moral judgment and decision making. Cognition, 108, 381-417.

Crockett, M. J., & Clark, L., & Hauser, M. D., & Robbins, T. W. (2010). Serotonin selectively influences moral judgment and behavior through effects on harm aversion. Proceedings of the National Academy of Sciences, 107(40), 17433-17438.

Greene, J. D., Nystrom, L. E., Engell, A. D., Darley, J. M., Cohen, J. D. (2004). The neural bases of cognitive conflict and control in moral judgment. Neuron, 44, 389-400.

Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., & Cohen, J. D. (2001). An fMRI investigation of emotional engagement in moral Judgment. Science, 293, 2105-2110.

Hannikainen, I., Miller, R. M., & Cushman, F. A. (2014). Act versus Impact: Conservatives and liberals exhibit different structural emphases in moral judgment, in prep.

Koenigs, M., Young, L., Adolphs, R., Tranel, D., Cushman, F. A., Hauser, M. D., & Damasio, A. (2007). Damage to ventromedial prefrontal cortex increases utilitarian moral judgments. Nature, 446, 908-911.

Miller, R. M. & Cushman, F. A. (2013). Aversive for me, wrong for you: First-person behavioral aversions underlie the moral condemnation of harm. Social and Personality Psychology Compass, 7(10), 707-718.

Moore, A. B., Clark, B. A., & Kane, M. J. (2008). Who shalt not kill? Individual differences in working memory capacity, executive control, and moral judgment. Psychological Science, 19, 549-557.

Paxton, J. M., Ungar, L., & Greene, J. D. (2012). Reflection and reasoning in moral judgment. Cognitive Science, 36(1), 163-177.

Youssef, F. F., Dookeeram, K., Basdeo, V., Francis, E., Doman, M., Mamed, D., Maloo, S., Degannes, J., Dobo, L., Ditshotlo, P., & Legall, G. (2012). Stress alters personal moral decision making. Psychoneuroendocrinology, 37(4), 491-498.

Dr. Ken Broda-Bahm is a Senior Litigation Consultant for Persuasion Strategies and has provided research and strategic advice on several hundred cases across the country for the past sixteen years. He has trained and consulted in nineteen countries around the world, is a past President of the American Society of Trial Consultants, and blogs at Persuasive Litigator, an ABA Journal Blawg 100 honoree.

Another Reason for Plaintiffs to Get Past The Victim Focus…

The center of the story in a plaintiff’s case has been shifting in recent years. This shift is due to better understanding of the ways jurors make moral judgments in evaluating parties. The transition is also thanks in part to fresh perspectives like the Reptile encouraging trial lawyers to tell stories that induce jurors to move beyond empathy, responding instead to more direct and personally-relevant appeals. The findings shared in this article by Hannikainen, Miller, and Cushman add to this discussion and, in my view, buttress the reasons for plaintiffs to move beyond traditional but dated ways of framing their cases.

The focus of the present article is on the ways research participants decide a given behavior is harmful. The team finds that participants’ own projected aversion to an action can be a better predictor of moral condemnation than the perceived consequences for the victim. Instead of simply assessing consequences, we also evaluate by testing whether we ourselves would feel bad engaging in those actions. The article ends with a brief discussion of the implications of this research for criminal law and the criminal jury. My goal in this brief comment is to extend this discussion by adding some thoughts on the message for civil plaintiffs and, by extension, for civil defendants as well. Ultimately, I think the study adds to the reasons plaintiffs have for embracing a relatively new way of framing cases, while suggesting some ways that defendants should adapt as well.

How to Be an ‘Old School’ Plaintiff

Conjure up a picture of the stereotypical plaintiff’s attorney addressing a jury, and the image will center on a few strong implicit features of the message:

· This case is about sympathy…

· The center and the starting point for this story is the plaintiffs themselves…

· We’re hoping that you’ll identify with and care about these victims, and that will motivate you to give a big award…

To Hannikainen, Miller, and Cushman, this focus jibes well with what has been the traditional perspective on moral evaluation, which “is often thought to derive from empathy toward the innocent victim”. The problem, as revealed in this study, is that this kind of simple empathy is neither the only nor the strongest motivator. The consequences for the plaintiff still matter, of course. That is your damages case. But it is not, it seems, the dominant factor in motivating the average juror to side with you. Instead, these studies suggest jurors are likely to focus at least as much on the defendant’s actions themselves and how justified jurors would feel in performing those actions themselves.

Those evaluations are made in a research context and should be tested in trial simulations. But it is reasonable to believe that the dynamic focusing on action-aversion would be as strong or even stronger in the courtroom. In a trial context, jurors are already primed these days to believe the messages of the tort reform movement: Juries are too easily moved by sympathy, too quick to translate injury to blame, and too open to the ideas of deep pockets and jackpot justice. Not wanting to be one of those juries, the panel is apt to react negatively if the message comes too early or too strongly that this case is about an injury and the sympathy they are supposed to feel about that injury. In the case presentation process, sympathy and a victim-focus backfires. Adding in the current study results, there are more reasons to say “No” to the old school style of presentation.

Transitioning to the ‘New School’ Plaintiff

The article also helps in clarifying what the new school would be. The answer is to focus on the action from the perspective of the actor. Intuitively, plaintiffs might think, “I don’t want jurors to identify with the defendant” but a personal aversion to the action itself seems to be a crucial step in moral evaluation. When trying to decide whether a given action is blameless or blameworthy, evaluators don’t simply look at the consequences of the action. Instead, they imagine what it would be like to perform that action and do a gut check on whether it would feel right or not if they did.

In a trial context, the focus on the action itself puts the defendant in the center of the story. It is jurors’ contemplation of that action which will be the source of the aversion that would motivate a verdict. “When judging third-party infractions,” the team concludes, “we imagine what it would feel like to perform the behavior oneself. In some cases, this process yields a feeling of aversion that contributes to our judgment that the behavior was wrong. ‘If it feels bad to me to do it, it’s wrong for you to do it.’”

To encourage that style of evaluation, plaintiffs should set the stage for jurors to try on that decision for themselves. If, in their own idealized projection, jurors would believe that they never would have acted as the defendant did, and if that action would feel wrong even in the contemplation, then they’ll be primed to condemn it, and to believe that it had compensable consequences.

Of course, this has a “What Would You Do?” feel to it, so counsel couldn’t embrace that message directly without running afoul of the “Golden Rule” objection. But centering the story on the decision-making and adding personalization and concreteness to that part of the story can serve as an invitation for jurors to assume that role. In addition, in oral voir dire, attorneys will have a legitimate right to ask about relevant attitudes and experiences, and that inquiry can also get jurors started in thinking about the decision from their own perspective.

What This Means for Defendants

The traditional civil defense focus of pulling jurors back to the facts, the law, and the evidence works quite well against the ‘old school’ plaintiffs who are aiming for a sympathetic focus on the consequences alone. Against the ‘new school’ types who want the defendant’s actions and not the plaintiffs’ injuries in the center of the spotlight, defense counsel need a new emphasis. Most basically, that means telling a positive story about the defendant’s commitments, choices, and actions so that jurors contemplating those behaviors for themselves will feel justification and not aversion. Or defendants might take an initially aversive act and try to make it less aversive by desensitizing jurors to it, for example, through repetition as the defense was reported to have done with the police beating video in the Rodney King civil trial.

But one critical way in which defendants might benefit from the researcher’s findings on action aversion is by applying it back on the plaintiffs. We know from experience that jurors’ tendency to ask “What would I have done in their shoes?” cuts both ways, and can easily lead to critical scrutiny of the plaintiff’s own actions. In an employment case, for example, it is nearly inevitable that every employed juror will see themselves as a kind of expert on what should happen in the workplace. Applying an idealized standard, they’ll tell themselves that they would have worked harder, documented more, or complained sooner if they were in the plaintiff’s circumstances at work. In other words, that same tendency to feel aversion as they consider an action applies to all the actions under evaluation within the case story, and not just the defendant’s.


The bottom line is that as the research on moral judgment and a number of related fields continues to bloom, practical litigators need to keep pace. That might mean the death of the idea that there is a time-honored, tried-and-true way to try cases. Not only does every case differ, but new studies are continually improving our practical understanding of evaluation, decision-making, and persuasion. This article from Hannikainen, Miller, and Cushman is a good contribution.

Response to “If It Feels Bad to Me, It’s Wrong for You: The Role of Emotions in Evaluating Harmful Acts”

Alison K. Bennett, M.S., a Senior Litigation Consultant with Bloom Strategic Consulting, has accumulated extensive nationwide civil and criminal litigation consulting experience. Her specialties include witness communication training, jury research in the form of focus groups and mock trials, and jury selection.

This journal article purports to examine the mental machinery behind how humans decide when harmful behavior is justified by the benefits it brings about. The authors designed and conducted four experiments to study the issue and concluded that both action aversion – aversion to the act of harm itself, and outcome aversion – aversion based on empathy for the victim, could be viewed as independent effects to explain differences in moral judgments. This hypothesis is interesting and bears study, as it could have profound implications for trial strategy for both criminal and civil law, some of which were outlined in the paper.

On the positive side, the experiments were inventive and the authors’ hypothesis appears to be worthy of further research. So, as a preliminary study, this article merits further discussion and adds value. On the other hand, the conclusions of the first three experiments were difficult to connect to the hypothesis, which was to examine the mental processes people use to decide when harmful behavior is justified by the benefits it brings about. The scientific methodology, which was lightly discussed, bears questioning as well. The fourth experiment appeared to test the authors’ hypothesis, but the conclusion that “the effects… explained almost all the variability in people’s moral judgments” could be viewed as over-reaching. Another concern is the implication reached after all four experiments, that “when judging third-party infractions, we imagine what it would feel like to perform the behavior itself.” The basis for this conclusion or connection to the findings of any of the experiments escaped me.

This study reminds me of experiments that preceded the seminal 2012 study[1] of whether nonverbal information significantly increase the accuracy of people’s judgments of trustworthiness. Prior to the 2012 study, numerous studies had been conducted to explore and test the issue, but all were flawed because they could not accurately isolate judgments of nonverbal information from other effects, such as the likability of the people giving examples of the nonverbal cues. In the 2012 DeSteno study, the authors were able to isolate the effects for testing by using a robot programmed to act out different facial expressions and nonverbal body cues, which removed the issue of if the participants were influenced by people illustrating different facial expressions and nonverbal behavior or actions, or the nonverbal actions alone. It took a number of years to reach the best experiment design to test the hypothesis, and this study may be a good preliminary step in testing the intriguing hypothesis proposed by the authors. To that end, if this study leads to the study of the effect of action aversion versus outcome aversion on moral judgments, it will have served a very useful purpose.

[1] DeSteno et al. (2012) Detecting the trustworthiness of novel partners in economic exchange, Psychological Science XX (X) 1-8.