The 55mph speed limit on US roads: comments on Godwin and Kulash's analysis

There is probably no way to evaluate the impact of the 55 mph limit in 1974 objectively—the necessary data were never collected. Godwin and Kulash assume a current saving of 3000 lives per year. This number is not the objective result of some explicit procedure, but rather the judgement of two conscientious experts; it is not subject to replication by other experts. Furthermore, the quality of their estimate is limited by the quality of the historical studies they are using for input: those studies relied on ‘best guesses’ because they did not have the necessary basic data. No amount of heroic effort by Godwin and Kulash can overcome these inherent deficiencies.

It is almost as difficult to quantify the benefits of the 55 mph limit today. The regression results reported above find no statistically discernible relation between average speed and the fatality rate, though they do show a significant relationship between speed‐variance and the fatality rate. Given this result, we want to know the relationship between the 55 mph limit and speed variance. There is weak evidence that this critical relationship is negative—raising the speed limit a small amount might actually decrease the dispersion of highway speeds.

Whatever the benefits of the 55 limit, its costs are important. On the rural interstate highways it costs about 250 years of extra driving time to save one life. And the policing resources spent on enforcing the 55 mph limit would save more lives if they were allocated to other highway safety measures.