This article explores a new driver of social desirability bias in survey research: threat-inducing violent events. We focus on fatal terror attacks as a specific example, investigating their impact using cross-national analysis and natural/survey experiments.

We find that these attacks lead respondents to overreport past electoral participation – a classic measure of social desirability bias. This result highlights the need for researchers to consider survey timing relative to such threatening events.

Key findings suggest some previously reported increases in political engagement following violent conflict may be inflated, due partly to this heightened self-presentation pressure. The study underscores the importance of accounting for real-world violence when interpreting survey data.