Research has shown that people perceive themselves as less biased than others, and as better than average in many favorable characteristics. We suggest that these types of biased perceptions regarding intentions and behavior of others may directly affect people’s decisions. In the current research we focus on possible influences in the context of helping behavior. In four experiments we found that, people believe that others, compared to themselves, are less inclined to help and cooperate, are less aware of the number of bystanders and more influenced by the “proportion dominance” bias and by the “identifiable victim effect.” We demonstrate that these perceptions are naïve and unrealistic by showing that decisions from both self and others’ perspectives are equally biased. Finally, we show how the perspective from which a decision is made (self vs. others) may affect private as well as public decisions in ways that might not be in the best interest of the decision maker and the public.

When we make risky decisions for others, we tend to follow social norms about risks. This often results in making different decisions for others than we would make for ourselves in a similar situation (i.e., self-other discrepancies). In an experiment, we investigated self-other discrepancies in young adults’ decisions to purchase a vaccine against a sexually-transmitted virus for themselves or for another person (i.e., the target of the decision). When the target’s preferences were in line with social norms, surrogates showed large self-other discrepancies in line with these norms. When the target’s preferences were contrary to social norms, surrogates did not show self-other discrepancies in line with these preferences; instead they still followed social norms, F(1, 140) = 21.45, p < .001, ηp2 = .13. Surrogates with lower numeracy, F(2, 128) = 3.44, p = .035, ηp2 = .05, and higher empathy, F(2, 128) = 3.72, p = .027, ηp2 = .06, showed self-other discrepancies more in line with the target’s preferences, even when these were contrary to the norm. Surrogates whose own risk attitudes were contrary to social norms showed larger self-other discrepancies, F(1, 128) = 5.38, p = .022, ηp2 = .04. These results demonstrate that perceived social norms about risk can predict self-other discrepancies in risky decisions, even when the target’s preferences are known and at odds with the social norm. Further, the surrogates’ numeracy, empathy, and propensity to take risks influence the extent to which risky decisions for others resemble risky decisions for oneself.