Road congestion and incident duration

- We estimate the causal effect of incident duration on drivers' time losses through changes in non-recurrent road congestion.
- Incident duration has a strong positive, but concave, effect on non-recurrent congestion.
- The duration elasticity of non-recurrent congestion is about 0.35 implying that a one minute duration reduction generates a €57 gain per incident.
- At very congested locations, the benefit of reducing the incident duration by one minute is about €1200 per incident.
- Public policies that prioritize duration reductions at congested locations are more beneficial.

Non-recurrent congestion is frequently caused by accidents and other incidents. We estimate the causal effect of incident duration on drivers' time losses through changes in non-recurrent road congestion on Dutch highways. We demonstrate that incident duration has a strong positive, but concave, effect on non-recurrent congestion. The duration elasticity of non-recurrent congestion is about 0.35 implying that a one minute duration reduction generates a €57 gain per incident. We also show that at locations with high levels of recurrent congestion, non-recurrent congestion levels are considerably higher. At very congested locations, the benefit of reducing the incident duration by one minute is about €1200 per incident. Public policies that prioritize duration reductions at congested locations are therefore more beneficial.