Which Political Activities Are Caused by Education? Evidence from School Entry Exams

Research on education's effect on participation is split between those who argue that education is a ``universal solvent'' that causally increases participation and those who hold that education merely proxies for pre-adult differences, with no independent effects. We incorporate a calculus of participation into the education-participation nexus to predict education’s effects for political activities that vary in their costs and benefits. We test these predictions with quasi-random variation in education caused by entry exams into upper-level secondary schools in Switzerland and participation measures from an original survey of former students conducted 40 years later. Comparing former students who narrowly passed or failed their exam, we find that an additional year of education increases electoral and low-cost non-electoral participation. We find no effects on high-cost non-electoral participation, however. These findings suggest that, rather than being a universal solvent or a mere proxy, education's effect depends on the characteristics of political activities.