Download data and study materials from OSF
Linda J. Skitka
University of Illinois at Chicago
Sample size: 395
Field period: 1/11/2006-1/16/2006
The study examined whether established conflict resolution techniques, such as providing opportunities for voice, could generate tolerance of non-preferred conclusions to moral disagreements or if moral conviction about outcomes restricted voice effects on justice judgments and related perceptions. Perceptions of voice were manipulated successfully by varying whether people were people that their survey responses would be shared with policy makers and by providing them with an open-ended opportunity to express their views. Results indicated that moral convictions about policy decisions qualified the effect of voice on perceived procedural and outcome fairness of the Supreme Court. The effects of outcome preferences and moral convictions also determined whether people were willing to accept the Supreme Court decision as the final word on the issue, and these effects overwhelmed the effect of voice effect on decision acceptance. In short, moral conflict appears to be more difficult to manage than non-moral controversies.
The studies tested two competing hypotheses: (a) the voice hypothesis predicted that people should perceive situations to be fairer and be more willing to accept decisions when they have voice than when they do not. Moreover, the effect of outcome preferences should decrease when people have voice compared to when they do not, irrespective of whether people’s outcome preferences reflect their moral convictions; and (b) the moral mandate hypothesis predicted that people should perceive situations to be fairer and be more willing to accept decisions when an outcome is consistent rather than inconsistent with their moral mandates. Moreover, the effect of outcome preferences should increase and the effect of procedures should decrease when people hold their outcome preferences with strong rather than weak moral conviction.
The experiment was a 2 (Voice: high, low) X 2 (Imagined Outcome: support, oppose) X 2 (Topic: abortion, gun control) X 2 (Topic order: abortion first, gun control first) mixed design. Topic and order were within-subjects variables, the other variables were manipulated as between subjects variables.
Voice. Participants in the high voice condition received instructions that indicated that the comments and responses they provided on the survey would be shared with their congressional and senatorial representatives, the United States Supreme Court, and the President of the United States. To reinforce the voice manipulation, participants in the high voice condition also were asked, “Do you have any other comments about abortion [gun control] policy that you would like to share with legislators?” and given a text box for their responses. Participants in the low voice conditions were not told that their responses would be shared and were not given the open-ended opportunity to provide comments.
Imagined Outcome. Following the voice manipulation, participants read about two hypothetical Supreme Court decisions. For each decision, the decision and participants’ attitudes about the issue jointly determined the experimental outcome condition. For example, half of the pro-choice participants were randomly selected to read that the Supreme Court overturned Roe v. Wade, making abortion illegal in the U.S. (the decision inconsistent decision condition). The other half of the pro-choice participants read that the Supreme Court reaffirmed its support for Roe v. Wade, keeping abortion legal and putting it on an even stronger legal foundation than before (the decision consistent decision condition). Similarly, half of the pro-life participants were randomly selected to read that the Supreme Court overturned Roe v. Wade, making abortion illegal in the U.S. (the decision consistent decision condition). The other half of the pro-choice participants read that the Supreme Court reaffirmed its support for Roe v. Wade, keeping abortion legal and putting it on an even stronger legal foundation than before (the decision inconsistent decision condition). Similar procedures were used for the gun control decision. Decision consistency or inconsistency was constant across topics.
Procedural fairness. Participants were asked the extent that they agreed or disagreed with the statement: “The procedures that the U.S. Supreme Court uses to make policy decisions are fair.”
Outcome fairness. Participants were asked the extent that they agreed or disagreed with the statement: “The policy decision to [make abortion illegal/keep abortion legal; make it more difficult/easier for people to purchase handguns] would be a fair and just conclusion to the debate about this issue.”
Decision acceptance. Participants were asked the extent that they agreed or disagreed with the statement: “I could accept the Supreme Court decision to [make abortion illegal/keep abortion legal; make it more difficult/easier for people to purchase handguns] as the final word on the issue.” Participants responded on 7-point bipolar scales scored +3 to -3 with scale point labels that ranged from strongly agree to strongly disagree.
Despite multiple indications that we successfully manipulated perceived voice and also perceived procedural fairness, results were more consistent with the moral mandate than the voice hypothesis. Specifically, moral mandates about the outcome moderated voice effects on procedural and outcome fairness. Voice enhanced perceived procedural fairness for people without a moral mandate, but voice did not affect perceived procedural fairness for people with a moral mandate. Additionally, voice had no effect on perceived outcome fairness for people who had a moral mandate about the outcome; people with a moral mandate perceived non-preferred outcomes to be quite unfair and preferred outcomes to be quite fair, irrespective of levels of voice. The effects of outcome preferences, and in particular morally mandated outcome preferences, also determined whether people accepted the Supreme Court decision, and these effects appeared to overwhelm the effects of voice on decision acceptance.
In conclusion, moral controversies appear to be inherently more difficult to resolve than non-moral controversies. Standard conflict resolution techniques, such as providing increased opportunities for voice in the process, appear to be less effective when people attach moral significance to policy decisions. Instead of focusing on the long-term instrumental and interpersonal implications of procedures, people’s perceptions of procedural fairness, outcome fairness, and willingness to accept policy decisions, people with a moral investment in outcomes focus nearly exclusively on whether their preferred outcomes are achieved.
Participants’ attitudes and moral conviction about abortion and gun control were assessed at the beginning of the survey, prior to the voice manipulation.
Bauman, C. W., & Skitka, L. J. (2006). Testing the generalizability of the moral mandate effect under conditions of personal voice. Paper presented at the biannual meeting of the International Society for Justice Research, Berlin, Germany.
Skitka, L. J. (2006). Moral conviction: Another dimension of attitude strength or something more? Invited address at the annual meeting of the American Psychological Association, New Orleans, LA.
Skitka, L. J. (2006). Moral conviction: Another dimension of attitude strength or something more? Invited address at the annual attitudes pre-conference at the Society for Personality and Social Psychology, Palm Springs, CA.
Skitka, L. J. & Bauman, C. W. (2006). Is morality always an organizational good? Paper presented at the Fifth International Round Table on Innovations in Organizational Justice: Justice, Ethics, and Social Responsibility, Tucson, AZ.