This is the second installment in a two-part article on persuasion research and its potential lessons for mediators. The first part appeared in the Fall 2011 edition of this magazine.
You are a law-trained neutral mediating an employment discrimination dispute. The 59-year-old plaintiff alleges that she was passed over for a promotion in favor of a younger male and then fired six months later from her office manager position because of age and gender discrimination. The company denies the charges, claiming that the plaintiff was laid off because of economic exigencies caused by the recession. Although her prior work evaluations were all good and there have been several provable incidents of sexist workplace remarks by supervisors, the plaintiff’s case is anything but a slam dunk. For one thing, she made a scene in front of her co-workers when she was passed over for promotion, referring to her supervisors in ways that even she now concedes were “unprofessional.”
Both sides are represented by counsel at the mediation, which has been in process for more than six hours. After an initial settlement demand of $200,000, a lengthy and heated factual exchange and a wide variety of non-evaluative mediator interventions, the parties seem tired and angry and are still $45,000 apart. You think that the plaintiff and her lawyer are not being realistic about their settlement goals. Various attempts at persuasion, including an unfavorable evidentiary prediction, have not made them change from their most recent position. If anything, she and her lawyer seem to be getting less reasonable as the mediation progresses. You think she has a substantial chance of losing the case altogether if it is tried but that there is a good chance the case will settle quickly if she is willing to drop her demand by, say, $25,000.
You are mulling over alternative ways of impressing this upon her. One option is to try to evoke fear, saying, for example: “I think you have a substantial chance of losing at trial. Win or lose, the incident in which you ‘dissed’ your employers in front of your co-workers will almost certainly be rehashed in detail. They may try to portray you as a loose cannon, disloyal, not a team player. With so many people looking for work in this economy, your conduct – even if it was justified – may look bad to a jury. In a slow news cycle, this might even be the kind of story that a newspaper or local TV station would want to cover. And if that occurs, the negative publicity could really hurt your prospects of being re-employed – no matter what else happened at trial.”
Another approach is trying to tap into her possible guilt, for example: “I know you’re a single mom who has been struggling lately to raise your teenage son, first on your unemployment benefits and now with part-time work. How will you feel about how you fulfilled your obligations to him if – after a long wait for the trial – you lose your case, knowing that you had the chance to receive some decent money in mediation but turned it down?”
If you are uncomfortable with either of these options, you are not alone. Many mediation theorists and practitioners might recoil at the suggestion that a neutral would actively seek to evoke fear, guilt or other “negative” emotions to produce a shift in attitude. Perhaps you are concerned that one or both of these options could manipulate or even overwhelm the plaintiff’s decision process. Apart from these normative implications, there are more basic questions: Do such approaches to persuasion work? If so, how and in what circumstances?
For more than a half-century, primarily in the areas of public health and advertising, social scientists have studied the effectiveness of fear appeals in seeking to influence behavior change. Early researchers saw such appeals in purely emotional terms: the fear these messages aroused created unpleasant effects, which drove subjects to seek relief by accepting the message’s recommendation. But more recent public health research has found that fear appeals produce not only emotional but cognitive responses and that their effectiveness may be highly tied to, if not determined by, the thinking they engender.
According to modern theorists, fear appeals can be broken down into two parts: (a) the threat or danger facing the subject, and (b) the recommended action or solution that the subject can take to avoid the harm. Each of these, in turn, is made up of two additional components, seen from the subject’s perspective. A threat is judged by the subject on the basis of his perception of (1) its severity and (2) its relevance – his chances of suffering its effects. The recommended action is appraised in terms of two different kinds of efficacy: (1) the likelihood that it will succeed in eliminating the threat and (2) the subject’s own ability to carry out that action.
A typical experiment might manipulate these four variables, placing each subject in one of eight experimental conditions. A 1987 breast cancer-based study is illustrative. College-age female subjects were first given either a high-threat essay (coupling dramatic written and photographic descriptions of cancer, as well as the side effects of radical chemotherapy or mastectomy, with a focus on student susceptibility to the disease due to stress and poor diets) or low-threat essay (which had far less intense depictions of breast cancer and its effects and told subjects that students rarely contract the disease). In addition, the essay described either a high or low likelihood that the desired response (breast self-examination) would be effective in early cancer detection, and presented a high (breast exams are easy to do correctly) or low (breast exams are difficult to do correctly, lumps are hard to detect) self-efficacy message.
Studies of this kind suggest that the success of a fear-based appeal depends on how subjects process the fear-inducing information. When a subject assesses the threat as low (because it is not severe, she is not susceptible to it, or both), little fear is aroused and she has little motivation to process the message further or do anything in response. Thus, low-threat fear appeals are not persuasive. But when the subject’s perception of both components of the threat-vulnerability combination is strong, she becomes motivated to do something to protect herself.
What that something is depends on what she thinks about the effectiveness of the recommended action. If she believes in the efficacy of the recommended coping action (both generally and in terms of her own ability to carry it out), the subject is likely to seek to control the danger by carrying out the recommended action, i.e., she is persuaded. But if she does not see herself as able to carry out a response that will be effective, the subject is more likely to seek to control her fear through denial of the danger, avoidance of a decision, suspicion of the source, wishful thinking or other resistance. Overall, the literature tells us that appeals that generate the most fear can be the most effective, so long as they convey both serious problems and strong, feasible solutions.
The use of guilt as a persuasive tool is commonly found in efforts to induce greater volunteerism or charitable giving and to influence consumer purchasing. By directly or indirectly confronting audiences with discrepancies between their personal standards and their actual conduct, guilt appeals seek to trigger unpleasant feelings that will motivate people to seek relief by making amends for their self-perceived shortcomings.
Efforts to persuade by capitalizing on guilt take several forms. The most direct involves the presentation of a message drawing a subject’s attention to the inconsistency of her behavior with her own standards or ideals. Anticipated guilt can also serve as a persuasive device. Subjects are asked to forecast how they will feel in the future if their current action or inaction produces harmful effects or causes them to fall short of their own standards.
Like fear appeals, guilt-based persuasive messages have a “problem-solution” structure: (a) the guilt-inducing suggestion that the recipient’s conduct or inaction violates her personal norms or some social or moral principle, and (b) the recommended change in behavior that can make up for the lapse and thus reduce the guilty feeling.
The extent of research in this area is rather limited as compared to empirical work concerning fear appeals. The field’s one meta-analysis, based on approximately 30 years of guilt studies, summarizes the research as follows: as with fear appeals, more intense guilt messages arouse a greater amount of emotional response. But unlike fear appeals, guilt appeals can go too far: as they become more intense or explicit, they become dependably less persuasive.
Illustrative of the experimental methodology in this area is a 1995 study of advertisements aimed at encouraging working mothers to purchase dental floss for their children. Subjects in the study were given informational statements that attempted to stimulate varying degrees of guilt: low (“Keeping your child’s teeth clean and fresh...that is FLOSS-IT’s job!”), medium (“You shape your child’s dental health, so don’t let your family down.”) and high (“It’s YOUR responsibility to make sure that your kids have healthy teeth and gums...don’t make any mistakes...DO IT RIGHT!”)
The results of this study showed first that the amount of guilt felt was higher in response to moderate- and high-guilt appeals than to low-intensity messages; second, that as guilt content rose from moderate to more blatant, such appeals generated anger and negative source attributions, (e.g., “The company is trying to manipulate my attitudes and feelings”); and third, that purchase intent declined as guilt levels rose. Strong guilt messages are perceived as attacks on the self and efforts to limit one’s freedom, coming from sources not entitled to criticize one’s conduct. Psychological reactance – resentment, anger and a desire to lash out against the message and the messenger – is often the result. Similar outcomes have been reported in studies involving transgressions in close or intimate personal relationships.
Interestingly, the use of anticipated guilt – illustrated by the mediator’s second option in the employment discrimination case described above – produces different results. In one recent study, undergraduate student subjects were asked to get tested to join the bone marrow donor registry during National Bone Marrow Awareness Month. While all students were given the same written description of the diseases for which bone marrow transplants might be life-saving and of the simple procedure for getting tested, the experiment varied the intensity of its anticipated guilt component. One group was given a “naturalistic” appeal that underscored the seriousness of the blood diseases that could be treated by donation and urged subjects to join the donor registry, concluding, “You may save a life!” The remaining subjects were given a high anticipated-guilt message that added brief stories about two children with leukemia – one who died for lack of a transplant, the other who flourished after getting one – and a closing exhortation to “think about how bad you might feel if you decided not to help when it is so easy.”
Compared to those who read the naturalistic appeal, students who read the more intensive message reported a higher estimate of the guilt they would feel if they failed to volunteer and a higher level of intention to take the first step to becoming a donor. Regardless of their level of intensity, such efforts to persuade were not met with psychological reactance; subjects generally found nothing objectionable about them.
In terms of effectiveness, both fear and guilt appeals may share one limitation that is significant in at least some mediation settings: their impact may be temporary. Fear appeals appear not to have lasting impact, at least as measured by the extent to which subjects actually carry out their intended behavior change (e.g., to stop smoking) over time. The impact of guilt appeals may also be ephemeral: For example, when surveyed a week or so later, students in the bone marrow experiment who had taken no action toward becoming a donor reported feeling less actual guilt than they had anticipated they would.
Mediation Applications and Questions
While this research has its origins in other fields, its findings may raise important questions for mediation practice. As stated, rather than evoking purely emotional reactions, successful fear appeals appear to trigger thinking, both about the threat and the subject’s ability to avert it. If this empirical literature is to be credited, the thinking triggered by effective fear appeals apparently can act to neutralize defensive and destructive tendencies such as anger, overconfidence, shame or denial. Viewed in this way, well-executed fear appeals by mediators – usually contained in an evaluative message – may actually work to enhance participant self-determination, at least in some cases.
Applied to mediation, fear appeal research may also point to the wisdom of directness, if not bluntness, in describing the severity of the risks and consequences of non-settlement options. Pulling punches or speaking in shorthand in delivering an evaluation may produce insufficient belief in the seriousness of the threat to trigger real openness to change. Moreover, if the perceived efficacy of a recommended action is an essential determinant of whether a fear-arousing message is accepted, merely stressing the risks and noxious consequences of a trial (or other alternative to settling), without more is unlikely to suffice. Embedded in message resistance may be a party’s belief that the opponent will never agree to settle. Or a party may doubt his or her own ability to afford or sustain the commitments needed to resolve the dispute.
A successful effort to persuade by sowing doubt or fear may therefore need to unpack the reasons for a party’s resistance to the message and provide at least general guidance, if not assurance, as to what action or offer will produce a resolution. It thus may follow that a that a negative evaluation may not be persuasive unless conducted relatively late in a mediation and only after a fair bit of caucusing. Once a mediator knows that one side will likely accept a specific offer, assuring the (fearful) opposing party that by making that offer she can put the threat behind her should be a relatively easy task.
Both direct and anticipated guilt appeals also show persuasive promise in mediation, but direct appeals do pose unique skills challenges. Finding a level of intensity that arouses motivating guilt but avoids psychological reactance seems difficult under any circumstances. If overdone guilt messages typically trigger resentment and anger even in personal relationships, the challenges for outsiders (like mediators) to leverage such bad feelings would seem even more daunting. And if that mediator has pledged to remain neutral and impartial, any obvious “guilt trip” might seem particularly likely to fail.
In addition, if fear and guilt appeals do not reliably produce lasting change in attitudes and behaviors, what about settlements that are brought about through such persuasion but require long-term commitments? If, for example, a mother’s agreeing to increased and regular child access for a father she hates was in any substantial measure the result of fear or guilt, might the type of persuasion used affect her willingness to abide by its terms over time? For many mediators concerned about the stability of agreements reached with their assistance, this measure of the quality of message acceptance could be an important consideration.
Implications for the Mediator Role Debate
Taken as a whole, the empirical research suggests that both fear and guilt appeals, skillfully executed, can be effective in changing minds. But as noted earlier, this begs the normative question: Are such “negative” emotional appeals an appropriate exercise in mediator persuasion?
As political scientist Drew Westin recently observed, “‘feelings’ are millions of years older than the kind of conscious thought processes we call ‘reason,’ and they have been guiding behavior far longer.” Mediation is often a highly emotional process in which the parties’ feelings hold strong sway on their communication patterns and decision-making processes – sometimes in ways that do not serve them well. Mediators who engage in any form of evaluation, sowing doubt about a disputant’s position or negotiating stance, do so knowing at some level that this may produce anxiety or even fear. Mediators who try to help parties honestly confront their own past bad behavior and contributions to a dispute directly or through orchestration of role reversals and apologies understand that these interventions may bring to the fore feelings of regret, guilt and shame. All such interventions seem well within the mainstream of accepted mediator practice.
Why is this so? If one thinks closely about the so-called “negative” emotions of fear and guilt, they are not really negative at all. Fear, of course, is highly adaptive when it helps humans escape danger or minimize risk. The ability to empathize with those we hurt is at the root of conscience and is what enables us to act morally toward one another. In the context of mediation, a healthy dose of fear induced by the mediator may help disputants reconsider overly confident decisions that may not be in their long-term interest. A dollop of induced guilt may help disputants come to terms more fully with the negative effects of their behaviors on others, thereby developing greater objectivity about their situation. So long as fear and guilt appeals are not exaggerated by the mediator, both kinds of interventions can produce more fully considered decisions – the kinds of outcomes every neutral should want to endorse.
James H. Stark is professor of law and director of the Mediation Clinic at the University of Connecticut School of Law. He can be reached at email@example.com. Douglas N. Frenkel is Morris Shuster Practice Professor of Law and directs the Mediation Clinic at the University of Pennsylvania School of Law. He can be reached at firstname.lastname@example.org. Their video-integrated text, The Practice of Mediation (Aspen Law and Business 2d ed. 2012), is used widely in law school classrooms. The full article on which this is based will be published at 28 Ohio State J. on Disp. Resol. 26241 (2013).
 See generally Roy F. Baumeister et al., Personal Narratives about Guilt: Role in Action Control and Interpersonal Relationships, 17 Basic & Applied Soc. Psychol. 173 (1995); Jeffrey Rubin & Warren F. Shaffer, Some Interpersonal Effects of Imposing Guilt Versus Eliciting Altruism, 31 Counseling & Values 190 (1987).
 William DeJong & Lawrence Wallack, A Critical Perspective on the Drug Czar’s Antidrug Media Campaign, 4 J. Health Comm. 155 (1999). On the other hand, some studies show at least some promise of long-term behavior change. See, e.g., Ronald W. Rogers et al., An Expectancy-Value Theory Approach to the Long-Term Modification of Smoking Behavior, 34 J. Clinical Psychol. 562, 564 (1978).