Combating influence operations is a critical concern for civil society, governments, and social media platforms around the world. Empirical research on how influence operations can affect people and societies—by, for example, altering beliefs, changing voting behavior, or inspiring political violence—is limited. To assess what is known about the effects of influence operations and identify research gaps we examine 82 studies published between 1995 and 2020 that: (1) examined a specific population targeted by an influence operation; (2) compared measurable outcomes (behaviors or beliefs) of people exposed versus those who were not; and (3) used a research design that provides credible evidence on the causal effect of the influence operation. The selected studies covered a range of influence operations, from political disinformation, state propaganda, and health misinformation.

This literature provides strong evidence that long-term campaigns on mass media have measurable effects on beliefs and consequential behaviors such as voting and risk taking combat. And there is some evidence that social media activity by exceptionally influential individuals and organizations can stoke low-level violence. But overall the literature still provides relatively little evidence on the potential impact of influence campaigns on social media.