Three essays on unethical behavior : the role of generalized reciprocity, discrimination and norms

Abstract

Understanding human behavior in its entire complexity is an ambitious if not impossible challenge. It is however possible to study particular aspects of human behavior through experiments that allow us to isolate specific facets in the decision-making process, ultimately leading to a better understanding of human behavior as a whole. This thesis covers three experimental articles on unethical economic behavior and sheds light on the motives and circumstances that lead individuals to engage in these activities. Clearly, unethical behavior in all its different manifestations can pose great risk to society – both at the large (e.g. corporate tax evasion) and small (e.g. shoplifting) scale – making it a relevant topic to be studied in economic research. Trying to understand unethical behavior through the lenses of traditional economic theory is problematic. Standard economic theory usually oversimplifies decision-making by assuming that people always take rational choices in a selfish-manner, irrespective of any normative considerations. However, this one-sided perspective on human behavior does not fully live up to what we observe and experience in daily life. Behavioral economic research enriched our understanding by acknowledging the relevance of apparent unselfish choices. Inspired by insights derived from behavioral science, economic experiments show that other-regarding preferences play a decisive role in the decision-making process (Cooper & Kagel, 2016). We observe that people deliberately repay kindness in a reciprocal manner and willingly help those in need as they donate money to charity. One way to conceptualize this apparent unselfishness is the idea that individuals tend to follow normative conventions, as they incorporate beliefs about what ought to be the right thing and act accordingly (Bicchieri, 2006). Overall, economic decision-making should be understood as a process in which both selfish (and often unethical) as well as other-regarding motives are balanced. Further, it is key to recognize this individual weighting of opposing motives not necessarily as a set of fixed preferences but as a dynamic procedure that can be shaped by situational factors or time. In particular, people may exploit given information to readjust the normative evaluation of their actions. For instance, it is a well-known empirical finding that people show a lower hesitation to litter in public when the environment around them is already filled with litter (Bateson et al. 2013). One may further think about the role of emotions in decision-making (Drouvelis & Grosskopf, 2016). Building on own experiences gathered in life, we know that acting in fierce anger or deep frustration can drive us towards decisions we would not take otherwise. These two simple examples illustrate nicely how malleable behavior is to the explicit situation we find ourselves in. In the following three chapters, I explore distinct decision-environments that open up the chance for decision-makers to justify – whether consciously or unconsciously – unethical behavior for their personal benefit. Chapter 2 of this thesis covers joint work with Sabrina Jeworrek and is based on the research article Jeworrek and Waibel (2021a). Here, we investigate whether individuals are willing to blindly pass on experienced unkindness to innocent others. In contrast to a standard gift exchange setting, in which unkindness would be directly reciprocated to the perpetrator, we build on a concept known as “generalized negative reciprocity” (Strang et al, 2016). This concept looks at the specific situation in which the original source of unkindness is not a feasible target for retaliation. Instead, it examines whether individuals are willing to displace their negative gift on to an unrelated, innocent party. Given the steep hierarchal structures that characterize many workplaces and the naturally high hesitation to reciprocate back to superiors, we causally test the relevance of generalized reciprocity in an artificial workplace setting. In a lab experiment, we assigned subjects into groups of three, promoting one to be the group leader. In the first stage of the experiment, the group leader had the chance to treat her subordinates in an unkind way by openly reassigning them to work on a tedious instead of a joyful task. In the second stage and independent from their former group leader, both group members worked on a new real effort task and competed for a monetary bonus. Similar to Charness et al. (2014), both competitors could freely choose to behave unethically by either manipulating their own or by sabotaging their opponent’s task score in order to increase their chance of winning. Our findings, however, do not support our main behavioral prediction. In comparison to the control group, subjects who experienced unkind leader treatment had a similar probability to cheat against their opponents in the subsequent tournament. With the help of a follow-up survey experiment, building on the norm-elicitation technique introduced by Krupka and Weber (2013), we confirm our lab findings. It shows that the moral appropriateness of cheating against someone who has suffered from the same superior mistreatment remained unaffected by the prior unkindness. However, in a second and slightly reframed workplace scenario, in which the counterpart was introduced as a new and unfamiliar coworker, the analysis shows that experienced mistreatment opened up some moral wiggle room to justify mild forms of cheating. Taken together, both experimental studies find limited support for the relevance of generalized negative reciprocity in our workplace setting while pronouncing the importance of the explicit social context for the emergence of the behavioral pattern. Chapter 3 also deals with the behavioral consequences of mistreatment on subsequent decisionmaking. Here, I openly provide participants with an excuse that can be utilized as a justification to disobey a clear request to comply with a given rule. Clearly, the willingness to comply with rules is an essential societal pillar, guaranteeing a peaceful cohabitation in life. Hence, understanding why and under which circumstances individuals refuse to follow rules is of great interest. One basic legal principle in modern societies is the convention of “rule equality”, ensuring that everyone plays by the same rules. In this chapter, I investigate whether an open violation of this convention can in turn undermine the willingness to comply. In the beginning of my online experiment, participants were first allocated into groups. In the main stage of the experiment, I introduced subjects to an unambiguous rule, prompting them to truthfully report private information. Inspired by the popular die rolling game (Fischbacher & Föllmi-Heusi, 2013), participants obtained their private information from three coin tosses executed in complete privacy. It was made clear that subjects who report three “heads” would receive a monetary price, while all other combinations resulted in zero payoff. Thus, everyone was free to disobey the rule and misreport in order to gain the monetary price. To causally test whether an unequal rule imposition within groups undermines the willingness to report truthfully, I randomly chose one participant in each group, who was publicly released from the original rule, while I still requested compliance from everyone else. In contradiction to my predictions, I find that unjustified rule discrimination had no impact on subsequent rule compliance. Furthermore, priorly induced group identities, aiming to increase the perception of equivalence among group members, did not change the results. Taken together, this project suggests that singular rule discrimination without severe monetary disadvantages seems to be insufficient to disrupt compliance across groups. In contrast to Chapter 2 and 3, which analyze the effect of prior mistreatment on the decision to behave unethically, Chapter 4 takes a different perspective. In the last research article, which is again a joint project with Sabrina Jeworrek, we examine how large societal changes can affect the willingness to behave pro- or antisocial. In this paper, which is based on a revised version of Jeworrek and Waibel (2021b), we build on the well-established finding that social norms are important but also fragile drivers for normative behavior in society (Kimbrough & Vostroknutov, 2016; Bicchieri et al, 2022). Exploiting the rigorous restrictions introduced during the Covid-19 pandemic, we examine whether social distancing and the resulting experience of extensive social isolation, influenced people’s perception of social norms and the willingness to comply with them. The paper draws on two online experiments. In the first, we elicited the normative beliefs on socially appropriate behavior in a Take-or-Give donation game, a game in which subjects could freely take (=unethical behavior) or give (=ethical behavior) money from or to a popular charity organization. In the second experiment, we observed actual decision-making within the same game but with new participants, allowing us to draw conclusions on the willingness to comply with the underlying social norms. To derive causal insights on the question whether persistent social isolation during the pandemic shifted norm perceptions and compliance, we used the priming method (Cohn & Maréchal, 2016) to make subjects’ personal lockdown and isolation memories salient. First, our results show that the normative expectations of appropriate behavior in the donation game did not change after recollecting individuals’ social isolation memories. The second experiment on norm compliance, however, revealed a clear difference. Here, our findings indicate that salient social isolation memories led to a decline in prosocial choices as primed participants took larger amounts away from charity in order to increase their personal payoff. Two additional information treatments, in which we simply reminded participants about the applying norms in the game, successfully increased prosocial behavior once again. Chapter 5 concludes this thesis and highlights the most important insights that can be drawn from the presented research articles.