Abstract: This paper proposes a parsimonious approach to estimate the effect of social interaction on stock market participation. Using data publicly available, we construct a sequence of measures of aggregated stock returns that are based on the same stock returns information and embed social interaction increasingly through different weighting functions. We show that the explanatory power over stock market participation is increasing in this sequence of aggregated stock returns measures. The effect is stronger when social interaction is stronger or social communication costs are lower, when the social information is positive and during bull markets. Our approach potentially is applicable to study the effects of social interactions in the aggregate in a wide range of contexts.