INTERPERSONAL RELATIONS AND GROUP PROCESSES Do Not Prime Hawks With Doves: The Interplay of Construct Activation and Consistency of Social Value Orientation on Cooperative Behavior Dirk Smeesters, Luk Warlop, and Eddy Van Avermaet Katholieke Universiteit Leuven Olivier Corneille Universite ´ catholique de Louvain at Louvain-la-Neuve and Belgian National Fund for Scientific Research Vincent Yzerbyt Universite ´ catholique de Louvain at Louvain-la-Neuve Low and high consistent pro-socials and pro-selfs were primed with neutral, morality, or might concepts in mixed-motive situations. The authors expected participants’ social value orientation to influence cooperative behavior among (a) high consistent individuals in all prime conditions and (b) low consistent individuals in the neutral prime condition only. The authors also expected the primes to influence cooperative behavior more among low than high consistent individuals. Four experiments using supra- liminal (Experiments 1, 2, and 4) or subliminal (Experiment 3) priming and 2-person (Experiments 1–3) or N-person (Experiment 4) social dilemmas partially supported these initial predictions. One intriguing exception was that morality primes reduced cooperation among high consistent pro-selfs. Experiments 2– 4 allowed testing for the potential role of expectations in shaping participants’ cooperative behavior. In interdependence dilemmas, individuals are faced with a con- flicting choice between the collective interest and self-interest. Individuals’ outcomes in these dilemmas do not only depend on their own choices but also on the choices of others. An individual is often tempted to make a noncooperative, self-interested choice because it yields higher personal outcomes than a cooperative, collectively interested choice, irrespective of what others might do. However, if all interested parties choose to pursue their self- interests, they are all worse off than if they had acted in a cooperative manner (Dawes, 1980). Therefore, mutual cooperative behavior is better for all parties than mutual defection. An abundant stream of research has demonstrated that social value orientation (i.e., a specific preference for various own– other outcome distributions; McClintock, 1972; Messick & McClintock, 1968) strongly influences cooperative behavior in mixed-motive situations (e.g., Kramer, McClintock, & Messick, 1986; Van Lange & Liebrand, 1989; Van Vugt, Meertens, & Van Lange, 1995). Some other research, however, has demonstrated that co- operative behavior could also be influenced by subtle situational cues such as primes (e.g., Hertel & Fiedler, 1994, 1998). These could even override the influence of social value orientation, which is often assumed to be a stable disposition (Dehue, Mc- Clintock, & Liebrand, 1993; Kuhlman, Camac, & Cunha, 1986; McClintock, 1972). According to Van Lange (2000; Van Lange, Agnew, Harinck, & Steemers, 1997), cooperative behavior in mixed-motive interdependence situations is best predicted by a Disposition  Situation interaction, such that the influence of social value orientation on cooperative behavior should be larger in situations lacking relevant situational cues than in situations with relevant, even subtle, situational cues (such as primes). Our objective was to extend and test this perspective. We predicted that the specific Disposition  Situation data pattern would depend on the consistency of one’s social value orientation. Hertel and Fiedler (1998) showed that individuals with a high Dirk Smeesters and Luk Warlop, Department of Applied Economics, Katholieke Universiteit Leuven, Leuven, Belgium; Eddy Van Avermaet, Department of Psychology, Katholieke Universiteit Leuven; Olivier Cor- neille, Department of Psychology, Universite ´ catholique de Louvain at Louvain-la-Neuve, Louvain-la-Neuve, Belgium, and Belgian National Fund for Scientific Research, Brussels, Belgium; Vincent Yzerbyt, Depart- ment of Psychology, Universite ´ catholique de Louvain at Louvain-la- Neuve. This research was supported by Belgian Office for Scientific, Technical, and Cultural Affairs Grant HL/DD/24 to Luk Warlop and Grant G.0260.02 from the Fund for Scientific Research—Flanders to Luk Warlop, Vincent Yzerbyt, and Olivier Corneille. We are grateful to Mario Pandelaere and Tim Smits for their help in conducting the experiments. This research was conducted while Dirk Smeesters was a doctoral student at the Katholieke Universiteit Leuven. This research was also presented at the Third Euro- pean Social Cognition Network, Houffalize, Belgium, September 2001. Correspondence concerning this article should be addressed to Dirk Smeesters, who is now at the Department of Marketing, Tilburg University, P.O. Box 90153, 5000 LE, Tilburg, the Netherlands. E-mail: d.smeesters@ uvt.nl Journal of Personality and Social Psychology, 2003, Vol. 84, No. 5, 972–987 Copyright 2003 by the American Psychological Association, Inc. 0022-3514/03/$12.00 DOI: 10.1037/0022-3514.84.5.972 972