Valuing statistical lives from observations of speed limits and driving behavior

The paper discusses how to derive empirical estimates of the value of a statistical life (VSL) from observations of highway driving speeds, and from how such speeds are affected by speed limits and penalties for speeding. When drivers optimize with respect to driving speeds, we discuss three alternative approaches. The first two rely on constructing drivers’ utility functions, and the last on revealed government preferences similar to that used by Ashenfelter and Greenstone (2002) (A-G). The two last approaches are based on observations of changed driving speeds when speed limits and speeding penalties change. When drivers are law obedient and adhere to speed limits only the A-G approach can be used. Their approach is however unrealistic in putting overly great demand on government information about VSL, and in addition provides upwardly biased average VSL estimates.