Clarification Regarding Rand et al. (2015) “Social Context and the Dynamics of Cooperative Choice”

Recent work using decontextualized economic games suggests that cooperation is a dynamic decision-making process: Automatic responses typically support cooperation on average, while deliberation leads to increased selfishness. Here, we performed two studies examining how these temporal effects generalize to games with richer social context cues. Study 1 found that time pressure increased cooperation to a similar extent in games played with in-group members and out-group members. Study 2 found that time pressure increased cooperation to a similar extent in games described as competitions and games described as collaborations. These results show that previous positive effects of time pressure on cooperation are not unique to neutrally framed games devoid of social context and are not driven by implicit assumptions of shared group membership or cooperative norms. In doing so, our findings provide further insight into the cognitive underpinnings of cooperative decision making. Copyright © 2014 John Wiley & Sons, Ltd. Additional supporting information may be found in the online version of this article at the publisher's web-site. key words cooperation; intuition; dual process; in-group bias; competition; framing; time pressure Cooperative dilemmas pit the costs associated with sharing one’s resources against the benefits that sharing creates for others. Given that cooperation is such an essential aspect of daily life (as well as a necessary condition for the continued survival of humankind), a great deal of research across psychology, economics, sociology, and biology has tried to understand and model how people make cooperation decisions (Apicella, Marlowe, Fowler, & Christakis, 2012; Axelrod, 1984; Bazerman, Magliozzi, & Neale, 1985; Bereby-Meyer & Roth, 2006; Bicchieri & Xiao, 2009; Crockett, 2009; Dawes, McTavish, & Shaklee, 1977; DeSteno, 2009; Fiedler, Glöckner, Nicklisch, & Dickert, 2013; Frank, Gilovich, & Regan, 1993; Fudenberg & Maskin, 1986; Fudenberg, Rand, & Dreber, 2012; Galinsky & Mussweiler, 2001; Gray, Ward, & Norton, 2014; Halevy, Weisel, & Bornstein, 2012; Locey & Rachlin, 2012; Pfeiffer, Tran, Krumme, & Rand, 2012; Rand & Nowak, 2013; Sanfey, Rilling, Aronson, Nystrom, & Cohen, 2003; Van den Bos, Vermunt, & Wilke, 1997; Zaki & Mitchell, 2011). Economic games have become a standard paradigm for exploring cooperation across fields because of their simple quantification of cooperativeness and their use of actual behavior rather than self-report measures or hypothetical scenarios (Camerer & Fehr, 2002). In these games, participants are given money and make choices between allocations that benefit themselves or benefit others. In the public goods game, for example, players decide how much money to keep versus contribute to a public pot, the contents of which is multiplied by some factor and divided equally among all of members of their group. Thus, the tension between individual and collective interests, which is at the heart of social dilemmas, is captured in a single incentivized decision. Here, we consider the cognitive dynamics of cooperation in these economic games. We do so using a dual-process perspective where decisions are conceptualized as resulting from the interaction between two different types of cognitive processes (Chaiken & Trope, 1999; Frederick, 2005; Kahneman, 2003, 2011; Miller & Cohen, 2001; Sloman, 1996; Stanovich & West, 1998): those that are fast, automatic, and intuitive and those that are slow, controlled, and reflective. How, then, do self-interested and other-regarding motivations map onto these different types of decision making? Is self-interest automatic, with cooperation requiring deliberation? Or, are first responses cooperative, with reflection leading toward selfishness? To shed light on this question, we focus on the effects of applying time constraints to economic game decisions. Time constraints are widely used in experimental psychology as a tool for investigating the role of intuition versus deliberation in decision making (Evans & Curtis-Holmes, 2005; Kahneman, 2011; Roberts & Newton, 2001; Suter & Hertwig, 2011; Wright, 1974). A key property of intuitive processing is that it is automatic and operates relatively quickly, whereas deliberative processing is typically slower and involves overriding automatic responses. Therefore, forcing people to decide quickly reduces their ability to reflect and gives them less opportunity to override their intuitive responses. Conversely, forcing people to stop and think has the opposite effect, allowing for more deliberation. Of course, the application of time pressure does not result in purely intuitive responding, nor does the enforcement of a delay result in purely reflective responding. Reflection may fail to override deeply held intuitions, and some subjects may engage in substantial reflection even under time pressure. Rather, by comparing time pressure and time *Correspondence to: David G. Rand, Department of Psychology, Yale University, 607-592-0218, 2 Hillhouse Ave, New Haven CT 06511, USA. E-mail: David.Rand@Yale.edu Copyright © 2014 John Wiley & Sons, Ltd. Journal of Behavioral Decision Making, J. Behav. Dec. Making (2014) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/bdm.1837 delay, we can reveal the dominant directions of the effects of intuitive versus reflective processing. Recent studies applying time constraints to economic games suggest that cooperative decision making is an inherently dynamic process: Time pressure sometimes increases cooperation in these games and sometimes has no effect but has not been found to systematically decrease cooperation (Rand, Greene, & Nowak, 2012; Rand & Kraft-Todd 2014; Rand et al., 2014; Tinghög et al., 2013). This pattern of results suggests that, on average, time pressure favors cooperation in economic games but that this effect is subject to various moderators. To explain this pattern of results, and to predict specific moderators, we have suggested the “social heuristics hypothesis” (SHH; Rand et al., 2014). The SHH posits that strategies that are advantageous (i.e., payoff maximizing) in daily life interactions become automatized as intuitions and get overgeneralized to less typical settings. The SHH therefore predicts that experiences outside the lab should moderate automatic responses (e.g., cooperation should be the default for people whose daily life interaction partners are trustworthy and who thus make cooperation profitable, but not for those who experience the world as untrustworthy). Consistent with this prediction, trust of daily life interaction partners has been shown to moderate the correlation between decision time and cooperation (Rand et al., 2012), as well as the effect of a time pressure manipulation (Rand & Kraft-Todd 2014). Additionally, the SHH predicts that prior exposure to lab paradigms should undermine the overgeneralization that drives automatic cooperation by giving participants a chance to develop new defaults for these games (or learn not to trust their intuitions in these settings). Consistent with this second prediction, selfreported level of previous experience with economic games moderates the effect of time pressure on cooperation (Rand & Kraft-Todd 2014; Rand et al., 2014), and a longitudinal analysis shows the remodeling of time-pressured responses over 2 years as economic game experiments became increasingly common in a particular subject pool (Rand et al., 2014). Finally, direct support for the SHH comes from experiments where participants are immersed in laboratory environments that favor either cooperation or noncooperation, establishing behavioral patterns that spill over into subsequent one-shot anonymous interactions, but only among subjects who rely on heuristics (Peysakhovich & Rand, 2013). A separate line of research suggests that cooperation levels in economic games can be influenced by subtle changes in social context. For example, changing the name of an identical prisoner’s dilemma payoff structure from the “community game” to the “wall street game” can substantially reduce cooperation (Dreber, Ellingsen, Johannesson & Rand, 2013; Liberman, Samuels & Ross, 2004). Similarly, describing the game as a competition reduces cooperation related to a neutrally framed baseline or a game described as a collaboration (Engel & Rand, 2014). Framing the cooperative decision as a “contribution to a public good” versus “extraction from a common resource” also effects cooperation levels (Brewer & Kramer, 1986; Fleishman, 1988) and priming the concept of money reducing helping (Vohs, Mead & Goode, 2006). In addition to this type of cue regarding the social environment, information about one’s interaction partner(s) also influences cooperation. For example, people are more prosocial toward those with whom they are more similar, such as sharing incidental similarities like birthdays or first names (Burger, Messian, Patel, del Prado & Anderson, 2004) or being members of the same group (Dunham, Baron & Banaji, 2008). The latter is true both for membership in trivial groups, such as those preferring the paintings of Klee or Kandinsky (Tajfel, Billig & Flament, 1971; Yamagishi & Mifune, 2008), and for more meaningful groupings, such as political affiliation (Fowler & Kam, 2007; Rand et al., 2009) or ethnicity (Whitt & Wilson, 2007). This susceptibility of cooperative decision making to social context raises interesting questions regarding previous results on time pressure in economic games. Were these earlier findings the result of participants implicitly assuming that partners in the game were in-group members or that the game was one where cooperation was expected of them? More generally, will time pressure continue to favor cooperation in social contexts that are unfriendly to cooperation (i.e., where overall cooperation rates are lower)? Here, we address these questions by crossing a time constraint man