Education and Political Participation

Narrative synthesis of the causal-inference literature on education and political participation

Adults and adolescents across the reviewed literature, mostly US and Western Europe.

Annual Review synthesis organised around three theoretical models of how education might cause participation, then a survey of the causal-inference literature. Graded C as a narrative review rather than a systematic one, but it is the most recent and most methodologically careful synthesis available and the best single entry point for the causal debate. READ DEPTH CAVEAT: read at abstract level via the publisher record and repository listings; specific effect sizes and study-by-study judgments were not extracted.

Three conclusions matter for this cluster. First, the causal-inference literature is genuinely mixed: studies using similar methods and similar data reach different conclusions about whether education causes participation, so the confident textbook claim that education causes voting is not supported by the identification literature. Second, on civic education specifically, the QUANTITY of civic education matters little for political participation. Third, HOW civics is taught matters more than how much - active-learning strategies show promise, especially for historically marginalised students. That last point should be read against Persson et al. 2020, a well-powered cluster-randomized test of exactly that hypothesis which found nothing, and against Jung and Gopalan 2024; the 'active learning works' claim rests largely on observational open-classroom-climate research.

Medium, inherited. The review's own point is that the underlying causal estimates are unstable; the observational civic-education studies it summarises on the 'how it is taught' question carry the standard selection and shared-family confounds that this archive's premise flags.

Its central observation IS about replication: studies using similar methods on similar data reach different conclusions about whether education causes participation. That is a statement that this literature does not replicate itself, and it is consistent with what the primary sources in this cluster show (US positive, Scandinavia null, draft-lottery null, twins mixed by country).