A Service of Leibniz-Informationszentrum econstor Wirtschaft Leibniz Information Centre Make Your Publications Visible. zbw for Economics Goeschl, Timo; Lohse, Johannes Working Paper Cooperation in Public Good Games. Calculated or Confused? Discussion Paper Series, No. 626 Provided in Cooperation with: Alfred Weber Institute, Department of Economics, University of Heidelberg Suggested Citation: Goeschl, Timo; Lohse, Johannes (2016) : Cooperation in Public Good Games. Calculated or Confused?, Discussion Paper Series, No. 626, University of Heidelberg, Department of Economics, Heidelberg, http://dx.doi.org/10.11588/heidok.00022247 This Version is available at: http://hdl.handle.net/10419/162969 Standard-Nutzungsbedingungen: Terms of use: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Documents in EconStor may be saved and copied for your Zwecken und zum Privatgebrauch gespeichert und kopiert werden. personal and scholarly purposes. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle You are not to copy documents for public or commercial Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich purposes, to exhibit the documents publicly, to make them machen, vertreiben oder anderweitig nutzen. publicly available on the internet, or to distribute or otherwise use the documents in public. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, If the documents have been made available under an Open gelten abweichend von diesen Nutzungsbedingungen die in der dort Content Licence (especially Creative Commons Licences), you genannten Lizenz gewährten Nutzungsrechte. may exercise further usage rights as specified in the indicated licence. www.econstor.eu University of Heidelberg Department of Economics Discussion Paper Series No. 626 Cooperation in Public Good Games. Calculated or Confused? Timo Goeschl and Johannes Lohse November 2016 Cooperation in Public Good Games. Calculated or Confused? Timo Goeschla Johannes Lohseb University of Heidelberg University of Birmingham November 24, 2016 Abstract Recent experiments suggest that contribution decisions in a public goods game (PGG) are more likely to be cooperative if based on intuition rather than reflection. This paper (i) reinvestigates the behavioral impact of so-called cognitive style in the PGG; and (ii) connects it with an earlier literature on the role of cognitive failure (confusion). This is motivated by the possibility that the method of time pressure, commonly used to identify cognitive style, invites confusion as a confounding factor. Two channels for such confounds are identified and experimentally tested: A heterogeneous treatment effect of time pressure depending on subject's confusion status and a direct impact of time pressure on subjects' likelihood of being confused. Our reinvestigation on the behavioral impact of time pressure confirms that cognitive style matters, but that deliberation rather than intuition drives cooperation. The confounding effect of confusion is not found to be direct, but to oper- ate through a heterogeneous treatment effect. Time pressure selectively reduces average contributions among those subjects whose contributions can confidently be interpreted as cooperative rather than confused. aEmail: [email protected]. Postal address: Department of Economics, Bergheimer Str. 20, 69115 Heidelberg, Germany. bCorresponding Author. Email: [email protected] 1 1 Introduction Economists' understanding of the drivers that underpin cooperation in the public goods game (PGG) has become increasingly sophisticated over the last twenty years (Ledyard, 1995; Chaud- huri, 2011; Vesterlund, 2014). Yet, despite the significant progress, it remains far from exhaustive and not without conflicting findings. As a result, researchers have adopted new conceptual ap- proaches, some borrowed from psychology or evolutionary biology, as points of departure for novel experimental designs and new interpretations of existing experimental evidence. Among the most recent approaches, cognitive style has attracted significant attention as a possible cog- nitive driver of behavior in the PGG. Cognitive style refers to the type of mental processes that individuals draw on in order to come to a decision in a given choice situation. In economics, researchers have started to adopt so-called `dual-system' models, which distinguish between intuition and deliberation as the two main cognitive styles (Loewenstein and O' Donoghue, 2007; Evans, 2008; Kahneman, 2011; Dreber et al., 2014).1 In the context of social dilemmas, economists' recent interest in cognitive style as a driver of cooperation has at least three rea- sons. One is the potential of dual-system theories to provide additional building blocks for modeling cooperative behavior. A second is the possibility of designing decision environments that enhance cooperation once the links between cognitive style, cooperative behavior, and the situational determinants of cognitive style are better understood. A third reason is that differ- ences in cognitive style could perhaps explain conflicting experimental findings that challenge existing outcome-based models of other-regarding preferences (Dreber et al., 2014).2 Cognitive factors have at least once before attracted the interest of economists studying subjects' behavior in the PGG. In his seminal paper, Andreoni (1995) showed that a significant share of behavior in the PGG that appeared to be cooperation could in fact be traced back to cognitive failure: Confusion about core elements of the strategic situation was common among subjects in the PPG. Subsequent research has been able to confirm the important role of confusion for explaining positive contributions. It has also demonstrated that repeated experiences and learning can attenuate its impact over time (Houser and Kurzban, 2002; Ferraro and Vossler, 2010; Burton-Chellew and West, 2013; Bayer et al., 2013; Burton-Chellew et al., 2016). In this paper, we connect the recent literature on cognitive style as an explanation of cooperative behavior in the PPG with the earlier literature on the role of cognitive failure among subjects playing that game form. These literatures not only share a common conceptual concern with cognitive factors in cooperation problems. More importantly, cognitive style and confusion are connected through the very methods commonly used to identify cognitive style. Because the cognitive processes involved in choice are not directly observable, researchers are beginning to collect supplementary non-choice data (Schotter, 2008; Schotter and Trevino, 2014), especially 1Apart from cooperation in the PGG, these models and the associated experiments have also been applied to other domains such as risky decisions (e.g., Kocher et al., 2013) or inter-temporal choice (e.g., Benjamin et al., 2013). 2In their paper, Dreber et al. (2014) describe a dual-self model of pro-social behavior that could reconcile some of these conflicting findings. For instance, standard social preference models cannot explain why a significant number of individuals chooses to avoid being asked to give, when there is a possibility to do so (Dana et al., 2007; Andreoni et al., 2011; DellaVigna et al., 2012). This behavior, however, would be consistent with a decision maker's deliberate choice to avoid the temptation of impulsive giving (Vesterlund, 2014; Dreber et al., 2014). Similarly, the difficulty to generalize from some lab results to other regarding behavior in the field (Levitt and List, 2007) could be due to the higher level of deliberation induced by the unfamiliar experimental environment. 2 on response times (Rubinstein, 2007; Spiliopoulos and Ortmann, 2014), or are putting subjects under time pressure, thereby exogenously influencing their cognitive style (Rand et al., 2012). Those and similar methods have been used separately or jointly in studies investigating the role of cognitive style in explaining outcomes of economic experiments in general and of social dilemma experiments in particular (Piovesan and Wengstr¨om,2009; Hauge et al., 2009; Fiedler et al., 2013b; Ubeda, 2014; Grossman et al., 2014; Lohse et al., 2014; Corgnet et al., 2015; Achtziger et al., 2015). A highly cited example is a series of one-shot PGG experiments (Rand et al., 2012, 2014) in which the researchers found that participants deciding under time pressure contributed more, on average, than participants deciding in a time delay condition. Their interpretation of this evidence as proof of a causal link between intuition and cooperation and, more generally, as support for applying a dual-self model to cooperation problems generated significant attention in the literature (G¨achter, 2012). If decision speed exclusively correlates with cognitive style, but is unconnected with the presence of cognitive failure, evidence from time-pressure experiments can be safely interpreted without reference to economists' earlier insights on the role of confusion in PGG. But there are at least two channels through which cognitive failure could conceivably confound the inference from time pressure experiments. The first channel is a heterogeneous treatment effect of time pressure on subjects depending on whether they are confused or unconfused about material aspects of the choice when they come to the decision situation.3 For instance, some subjects might simply have misread the instructions before they reach the decision screen and hence do not understand that there is a