Persuasion under the Threat of Verification

This paper studies persuasion when public information is costly and receivers can verify the state before acting. Verification may overturn acceptance after favorable public news, but it may also induce acceptance after unfavorable news. We characterize how these two effects shape the sender's optimal experiment. Cheaper verification can lead the sender to disclose less and rely more on receivers' own discoveries, thereby increasing its payoff. With heterogeneous decision thresholds, we identify exact conditions for the disclosure response and show why it depends on the distribution of preferences. In a separate parameter region, cheaper verification increases inspection but reduces every threshold type's final information and welfare; the sender also loses. We then allow receivers to share their findings. Cheaper inspection and wider sharing can move optimal public disclosure in opposite directions, even when both benefit the sender. Wider sharing can also reduce aggregate final information while improving every threshold type's welfare. The results distinguish the effects of verification on public disclosure, information ultimately obtained, and the cost of obtaining it.