This report shows that Daylight Saving Time (DST) reduces the number of persons killed in motor vehicle fatal traffic accidents by about one percent. This estimate is based on a spectral (Fourier) analysis of these fatalities which utilizes a filtering technique to identify the part of the fatality frequency spectrum which is sensitive to DST while suppressing all other frequencies. To establish a cause/effect relationship between DST and changes in the filtered fatality time series, the changes are measured in two ways: (1) Across DST transitions and (2) for corresponding dates in years with and without DST. Certain statistical criteria are then applied to these measurements in order to confirm the existence of a DST effect on accidents and fatalities.
Abstract