Abstract:
Do salient threats make people take more or less risk? I study this question using wartime air alerts in Ukraine, a recurring shock that is threatening, disruptive, and often occurs at night. Combining time-stamped alert records with daily administrative traffic-accident data from Lviv, I find that each additional hour of overnight alert exposure is associated with about 10 percent fewer next-day accidents. The decline is concentrated in afternoon and evening travel, weekends, central areas, and accident categories linked to risky driving, while fatigue-related accidents do not increase. These patterns are consistent with behavioural risk avoidance: after a salient threat, people appear to reduce or retime discretionary travel and drive more cautiously. Evidence from thirteen other Ukrainian cities suggests that this response is strongest where alerts are relatively rare and weaker where alerts are routine. Salient threats can therefore reduce risky behaviour, but the response appears to depend on continued salience.
