Pedestrian Fatalities and Daylight Saving Time

This graph, reproduced from the January 1995 edition of the American Journal of Public Health, shows the dramatic effect Daylight Saving Time (DST) has on the safety of pedestrians. Illustrated are the number of fatal pedestrian crashes in the US during the hour which ends at the time of sunset on the last Saturday of October, in the 13 weeks before and 9 weeks after the end of DST.
With DST, the average number of pedestrian fatalities is 25 or less; as soon as DST ends, pedestrian fatalities quadruple to over 100! This effect is so large, and over such a long period of time, that it almost certainly swamps any transient effect of tiredness just after DST is introduced.
Daylight, almost certainly because of improved visibility, had a substantial impact on the likelihood of pedestrians being killed on the roads. While the effect would be reversed in the mornings, very few pedestrians are on the roads at that time, so the reduction in fatalities from additional daylight in the evenings considerably exceeds the small increase from less daylight early in the morning.

More than 900 lives could be saved by extended DST
The published research predicted that if DST had been retained through the years of the investigation (1987 to 1991), there would have been a total of 727 fewer fatal pedestrian crashes and 174 fewer crashes fatal to vehicle occupants. Similar conclusions have been obtained from a number of independent studies in the US and UK.

Additional published research showing a reduction in road fatalities from extending DST
Broughton J, Sedman RJ. The potential effects on road casualties of double British Summer Time. UK Transport and Road Research Laboratory, research report 228, 1989.
Meyerhoff, NJ. The influence of Daylight Saving Time upon motor vehicle fatal traffic accidents. Accident Analysis and Prevention: 10: 207-221, 1978.
Joksch HC, Wuerdemann H. The impact of year-round Daylight Saving Time upon traffic deaths and injuries. Center for the environment and Man Inc.; CEM report 4166-506, 1974.
British Standard Time and Road Casualties. Transport and Road Research Laboratory, LF213, 1970.