Morality Games 307

Morality Games 307 Fig. 5 The Repeated Prisoner’s Dilemma. Two players play a Prisoner’s Dilemma. They each observe the other’s action, then, with probability 6, play another Prisoner’s Dilemma against the same opponent, etc. Technically, for the second and third point, what is needed is common knowI- edge that a player had an opportunity to cooperate or of the more effective means of cooperation. If observers were to know one purposely chose to defect or chose the less cooperative act, but they do not know that others know this, then observers think others will think punishment is not warranted, and observers will not punish. The argument is analogous to the discussion of higher-order beliefs in the omis- sion—commission subsection and formalized in Dalkiran et al. (2012) and Hoffman et al. (2015). Interpreting the Quirks of Altruism Below we discuss some of the quirky features of altruism identified by economists and psychologists. In each case, we will argue that these features might be puzzling, but not when viewed through the lens of the above model: Insensitivity to Effectiveness. We are surprisingly insensitive to the impact of our charitable contributions. We vote because we “want to be a part of the democratic sys- tem,” or we “want to make a difference,” despite the fact that our likelihood of swinging an election (even in a swing state) is smaller than our likelihood of being struck by lightning (Gelman, Silver, & Edlin, 2012). Why is our desire to “make a difference” or “be a part of the system’ immune to the actual difference we are making? Our chari- table contributions or volunteer efforts suffer from the same insensitivity. Why does anyone give money or volunteer time to Habitat for Humanity? The agency flies high eamers who have never held a hammer halfway across the world to build houses that would be substantially more cheaply built by local experts funded by the high earners. Experimental evidence demonstrates our insensitivity: Experimental subjects are will- ing to pay the same amount to save 2000, 20,000, or 200,000 birds (Desvousges et al., 2010). Likewise, when donors are told their donations will be matched, tripled, or quadrupled, they donated identical amounts (Karlan & List, 2006). Why do we give so much, but do not ensure our gifts have a large impact? The explanation follows directly from the above model: It is often the case that observers do not know which acts are effective and which are not and, certainly, this HOUSE_OVERSIGHT_015519