Narratives in the 2024 Moldovan Presidential Election: Stance, Coordination, and Influence

Online political discourse spreads across several platforms at once, which makes it hard to tell genuine peer influence apart from coordinated amplification and from users who simply resemble each other. This work studies the discourse surrounding the 2024 Moldovan presidential election and separates these effects. We characterize narratives and stance across platforms, detect amplification behavior, and estimate heterogeneous peer influence while accounting for latent homophily, combining causal proxy modeling with network exposure embeddings.