Social influence among people is widely understood to be a universal component of the human experience. However, studies of political behavior have generally approached social influence as specific to a type of behavior, such as voting, in a particular national context. There are good reasons to expect that social influence is observable across diverse behaviors and national contexts. In this study, we test this expectation using a two-wave panel survey of national samples in 19 countries. We employ autoregressive models that address some of the endogeneity challenges associated with attempts to measure social influence with survey designs. Our measure of social influence is predictive of diverse political behaviors in many countries with average effects comparable in size to important standard predictors of behavior.