Humans sometimes behave very selfishly, yet in other cases they prefer to help others. These “social preferences” in turn play a key role in many economic phenomena ranging from financial fraud to charitable donations. This dissertation presents a series of lab- and field-experiments about the drivers of social preferences, with a particular focus on bounded rationality, culture and beliefs. These experiments are conducted in a diverse set of populations, ranging from bank employees to entrepreneurs in a developing country to professionals in charitable organizations. The results indicate that humans differ considerably in their prosociality. Furthermore, well-intended “nudges” such as a group-level ethics training can be ineffective or e...