The Problem with All-or-Nothing Trust Games: What Others Choose Not to Do Matters in Trust-Based Exchange
22 Pages Posted: 29 Dec 2015 Last revised: 5 Jan 2016
Date Written: January 4, 2016
Many economic interactions are characterized by “all-or-nothing” action spaces that may limit the demonstrability of intended trust. We investigate whether restricting investment opportunities to all-or-nothing options affects the investment rate and propensity to reciprocate. We do this by manipulating the investor’s action space in two versions of the trust game. In the all-or-nothing game the investor can invest either $10 (all) or $0 (nothing), while in the continuous game the investor can invest any amount between $10 and $0. In both games, the trustee receives the tripled investment and then can return any amount to the investor. Results indicate that investments are higher in the all-or-nothing game than in the continuous game. However, higher investments in the all-or-nothing game do not lead to higher returns. To the contrary, conditional on $10 investments, on average trustees return less in the all-or-nothing game. Although the all-or-nothing action space results in greater wealth overall, it also appears to “backfire” for investors who do not benefit from the increased wealth. These results support the proposition that humans perceive intentions not only by evaluating what others do but also by evaluating what others choose not to do.
Keywords: trust game, demonstrability, intentions, reciprocity, experiment
JEL Classification: C72, C91
Suggested Citation: Suggested Citation