How Political Information Travels, and What It Does to Trust in Institutions

Most of what we know about how political information spreads comes from the places researchers can watch: television, newspapers, and the open social media platforms. Americans increasingly get their news somewhere else—in group chats and private messages, from podcasts, search results and AI assistants, and from people they know. These channels are not merely harder to study. Many are built so that no outsider can see what moves through them, and the rules written to govern online speech assume a visibility they do not have. This dissertation asks what follows. The first paper measures how far political conspiracy claims travel through private messaging, where neither platforms nor researchers can follow them. The second asks why American confidence in public institutions has fallen, and finds that what is usually described as a collapse of trust in science is largely something else: trust in agencies that changes hands with each election, while confidence in scientists themselves barely moves. The third maps the whole information environment by what can be counted and who can be held responsible, then simulates what becomes of a false claim, and of a correction, as audiences move into channels nobody can observe.