Trust, risk and betrayal

Trust at zero acquaintance: More a matter of respect than expectation of reward.

“…However, people were quite averse to giving the $5 to Person B when that person would flip a coin to determine the fate of the $20, despite such behavior being just as altruistic and efficient. Data from other studies shows this same pattern: Adding payoffs to other people beyond the self in a coin flip fails to inspire more gambling among participants (Bohnet & Zeckhauser, 2004;Schlösser, Fetchenhauer, & Dunning, 2013;Schlösser, Mensching, Dunning, & Fetchenhauer, 2014). Further, people are not influenced in the trust game by whether their interaction partner is poor, and thus needs the money, or is rich (Brülhart & Usunier, 2012).…”

Civic Capital as the Missing Link

“…Actual trust behavior, as measured for instance by the amount of money that a person would be willing to lend to an unknown individual, obviously depends both on the belief the lender has about the borrower's trustworthiness as well as on the lender's willingness to bear the risk that the borrower does not repay. When faced with -social risk‖ -that is the risk that a loss is caused by another person rather than nature -what matters is betrayal aversion (Bohnet and Zeckhauser, 2004) -that is the dislike for the risk of being cheated, not risk aversion. By using the German Socio-Economic Panel (which collects measures of trust, risk preferences and betrayal aversion), Fehr (2009) collected by Bohnet et al (2008) seems to suggest that risk and betrayal preferences do differ, though sample sizes are not large enough to draw strong conclusions (see also Naef et al 2008).…”

Complicity without connection or communication

“…Our starting point, when thinking about how to achieve this was Bohnet and Zeckhauser's (2004) experiment designed to isolate and measure betrayal (of trust) aversion. In their experiment the trusters and trustees in one treatment and the gamble choosers in the other, made eitheror decisions.…”

“…In their experiment the trusters and trustees in one treatment and the gamble choosers in the other, made eitheror decisions. This being the case, Bohnet and Zeckhauser (2004) could effectively hold all other things, except the existence of the trustees but including the trusters' ex ante subjective distributions of monetary payoffs conditional on their own decisions, constant by asking the trusters and gamble choosers for their "minimum acceptable probabilities (MAPs) of getting the good outcome such that they would prefer the chance to the sure payoff" (p.467). In this context, betrayal aversion manifests as trusters reporting higher MAPs relative to gamble choosers.…”

“…In the CG, play is simultaneous and each player has to choose one of six possible reports. Thus, we had to devise an alternative to Bohnet and Zeckhauser's (2004) MAPs approach to holding ex ante subjective distributions of monetary payoffs conditional on own decisions constant, while eliciting the players' decisions. In addition, like Bohnet, Greig, Herrman, and Zeckhauser (2008), we were interested in the specific effect of the presence of an active player in a particular role in a game, while all other social motivations are held constant, so we replaced the potential accomplice in the CG with a passive player in the NAc, rather than having no second player at all.…”