HOUSE_OVERSIGHT_015520.jpg

308
M. Hoffman et al.

usually is not commonly known. Thus, they will not reward or punish based on effectiveness, and we ourselves will not attend to effectiveness in equilibrium. This explanation suggests that if we want to increase efficacy of giving, we ought to focus on making sure donors’ friends and colleagues are aware of the efficacy of different options. In fact, this is perhaps more important than informing the donor of efficacy, since the donor will be motivated to uncover efficacy herself.

Magnitude of the Problem. We are surprisingly unaware of and unaffected by the magnitudes of the problems we contribute to solving. How many of those who participated in the recent ALS Ice Bucket Challenge have even the vaguest sense of the number of ALS victims? (Answer: about 1/100th the victims of heart disease.) How much happier would these individuals have been if the number of ALS victims were cut in half? Multiplied by 100? The same questions could be asked about AIDS or cleft lips. If we were actually motivated by our desire to rid the world of such afflictions as we often proclaim, then we would be happier if there were fewer afflicted individuals and less happy if there were more. But we are not even aware of these numbers, let alone affected by them. This suggests an alternative motivation than the one we proclaim.

On the other hand, if we give in order to gain social rewards, it does not matter whether the problem is large or small, provided others recognize it as a problem and the social norm is to give. If our learned or evolved preferences were drastically impacted by the magnitude of the crises, we would be sensitive to whether the problem was solved, perhaps motivating us to ensure that others solve it, which we would not get credit for, or perhaps motivating us to devote too much of our resources to solving it, beyond what we would actually get rewarded for.

Observability. There is overwhelming evidence that people give more when their gifts are observed. Much of this evidence comes from the lab, where it has been demonstrated a myriad of ways (e.g., Andreoni & Petrie, 2004; Bolton, Katok, & Ockenfels, 2005; List, Berrens, Bohara, & Kerkvliet, 2004). For instance, when participants play a public goods game in the laboratory for money, their contributions are higher when they are warned that one subject will have to announce to the room of other participants how much they contributed (List et al., 2004). However, evidence also comes from real-world settings, which find large effects in settings as diverse as blood donation (Lacetera & Macis, 2010), blackout prevention (Yoeli, Hoffman, Rand, & Nowak, 2013), and support for national parks (Alpizar, Carlsson, & Johansson-Stenman, 2008). In Switzerland, voting rates fell in small communities when voters were given the option to vote by mail (Funk, 2006), which makes it harder to tell who did not vote, even though it also makes it easier to vote. In fact, our willingness to give more when observed extends to subtle, subconscious cues of being observed: People give twice as much in dictator games when there are markings on the computer screen that vaguely represent eyes (Haley & Fessler, 2005), and they are more likely to pay for bagels in their office when the payment box has a picture of eyes above it (Bateson, Nettle, & Roberts, 2006).

These results should not surprise anyone who believes our pro-social tendencies are influenced by reputational concerns (though the magnitudes are surprisingly large).

HOUSE_OVERSIGHT_015520