Fear Appeals: Terror Only Works With an Exit

Howard Leventhal's tetanus experiments at Yale in 1965 are usually cited for the wrong reason. The finding people remember is that fear works. The finding that matters is what happened to the students who were frightened and given nothing to do.

Leventhal showed senior students one of two pamphlets about tetanus. The high-fear version had graphic photographs of infected wounds and a description of the disease that was accurate and appalling. The low-fear version conveyed the same information without the imagery. Both recommended vaccination, which was available free at the campus health service.

The high-fear group reported considerably stronger intentions to get vaccinated. Then Leventhal measured who actually turned up, and almost nobody did — in either group. Vaccination rates were about three percent.

He then added a component to both pamphlets: a map of the campus with the health service circled, and instructions to look at their timetable and pick a specific time to go. Vaccination rates rose to twenty-eight percent. The map did not increase fear. It did not add information about tetanus. Crucially, it raised rates in the low-fear condition too.

Kim Witte formalized what this implies in 1992 as the Extended Parallel Process Model, and it is the single most useful framework in this chapter. A threat message is evaluated on two dimensions: how bad and likely is the threat, and how effective and achievable is the recommended response. When perceived threat is high and perceived efficacy is also high, people enter danger control and act on the recommendation. When perceived threat is high and perceived efficacy is low, they enter fear control — and fear control does not mean doing nothing. It means managing the feeling rather than the danger, through denial, defensive avoidance, minimization, fatalism, or hostility toward the source of the message.

That last route is why frightening campaigns sometimes produce the opposite of their intent, and why people who are alarmed and helpless often turn on the messenger.

For anyone studying propaganda, the operational implication is stark. A population that is frightened and offered no efficacy path is not persuadable; it is defensive. A population that is frightened and offered one is highly persuadable, and it is not very particular about the quality of the path, because the psychological demand is for relief from the fear-control state. Supply a credible-sounding action, an identifiable cause, or a person to blame, and the audience will take it, because the alternative is remaining in a condition the mind is actively trying to escape.

This is the structure underneath a great deal of the material in Part 7 and Part 8 of this guide. Threat, plus one prescribed remedy, plus a monopoly on the remedy. The demagogue does not merely frighten people; anybody can frighten people. The demagogue frightens people and is the only one holding a map.

The countermeasure follows directly from the model and is unusually clean. When a message frightens you, separate the two claims it has bundled: is the threat real and as described, and is the proposed response the best available one. The bundling is the manipulation. A real threat can be paired with a terrible remedy, and the fear you feel about the first is not evidence about the second — but it will be spent on it unless you deliberately pull them apart.