Recent studies on turnout have provided convincing empirical evidence suggesting that voting can be habit forming. What exactly this means in theoretical terms has been scarcely investigated how-ever. Accordingly, there has been hardly any attempt to incorporate this theoretical claim within more established schemes about the determinants of turnout. Building on this research, the present study attempts to systematically examine the long-term implications of initial electoral experiences. To do so, we start by conceptualizing the logic of repeated voting behavior as a resource appearing during people’s life trajectories. We test this claim by comparing respondents who differ in nothing but the type of their first eligible election. Turnout in midterm elections is constantly lower than in Presidential elections. Thus, by distinguishing between the two we obtain some leverage which we use in order to identify the effect of early electoral records. The findings support this line of argu-ment, suggesting that becoming first of age during a Presidential election boosts the likelihood of voting in a midterm election by approximately 10 per cent.
