The differences in effects of social image by gender using risky dictator game experiments

• This study conducts risky dictator game experiments in which recipients' payoffs are probabilistically determined by dictators' allocations. • Because recipients cannot infer the allocation, they do not need to worry about their social image. • By informing the allocation to recipients, we can construct a situation in which dictators can care about their social image. • We measure the effects of social image on prosocial behavior by comparing the allocation of risky dictator games with and without social image. • This study observes gender differences in the effects of social image: females care about their social image, but males do not. Experimental evidence has revealed that females are more prosocial than males. However, we do not know much about what lies in differences in prosocial behavior in gender. The research question of this study is how effects of social image differ by gender and these effects can explain gender differences in prosocial behavior. Social image is a desire to be perceived as fair and is impure motivation behind prosocial behavior. Experimental studies developed various devices to extract social image and observed its existence in dictator game experiments. However, these methods are not suitable for our purpose because we need to measure the effects of social image, not just existence. This study conducted the risky dictator game in which dictators do not care about their social image because recipients cannot infer dictator's allocation. By adding social image, we prepare the two types of risky dictator games with and without social image. We measure social image based on differences in amount and probability of positive allocation in risky dictator game experiments with and without social image. This study observed differences in effects of social image by gender. We draw conclusion that social image is a cause for gender differences in prosocial behavior.