It is important to understand how speed impacts safety, because setting speed limits is primarily a road safety measure. While the laws of physics make it very clear that speed and crash severity are inextricably linked (i.e., severity increases geometrically as speed increases), there has been a good deal of controversy over the impact of speed on crash occurrence. This is primarily because the variety of road design and operating characteristics can obscure the precise relationship between speed and crash occurence. Numerous studies and research efforts on this topic that have presented conflicting results on this important relationship. However, the most recent and statistically robust research on speed and crash occurrence fairly definitively indicates that, all other factors being equal, increased speeds increase crash occurrence.⁷ The magnitude of the increase is dependent on the specifics of each case, with urban areas having the most pronounced relationship and controlled-access facilities the weakest.

One of the most statistically robust efforts to uncover the relationship between speed and safety was a meta-analysis conducted by the Norwegian Institute of Transport Economics.⁷ The information and conclusions from the meta-analysis form the basis for the statements made in this section.

For a given roadway type, there is a strong statistical relationship between speed and crash risk for speeds in the range of 15 mph to 75 mph (25 km/h to 120 km/h). When the mean speed of traffic is

educed, the number of crashes and the severity of injuries will almost always go down. When the mean speed of traffic increases, the number of crashes and the severity of injuries will usually increase. The relationship between mean travel speed and crash risk can be adequately described in terms of the following model:

(Extended Text Description: Equation: CMF = (V_a / V_b) ^ X )

CMF = Crash modification factor

V_a = Mean speed in the after condition

V_b = Mean speed in the before condition

X = 3.6 for fatal crash frequency

2.0 for injury crash frequency
1.0 for property-damage-only crash frequency
4.5 for fatalities
2.7 for personal injuries

The relationship between speed and crash risk can be modified to some extent by road environment, vehicle-related factors, and driver behavior. But, the effects of speed on crash risk are remarkably consistent across different contexts.

The above relationship between speed and crash risk is significantly different from the traditional U-shaped relationship that has defined much of the current North American thinking on speed limits and speed management. The U-shaped relationship (Solomon curve) between speed and crash risk can be questioned for two reasons:

1. The U-shape is generally expected to be an artifact of errors in the measurement of speed^8,9; and

2. There is a strong correlation between mean speed and speed variance, so it is difficult to separate the effects of mean speed and speed variance on crash risk.¹⁰

This discussion describes the relationship between travel speed and crash risk, but it does not necessarily reflect the relationship between speed limits and crash risk.

A change in the speed limit almost always changes the mean speed of traffic. However, the changes are not always proportional. For the most part, the change in the mean speed of traffic created by a change in speed limit is around 25 percent of the change in the speed limit.⁷ In other words, a speed limit increase or reduction of 6 mph (10 km/h) yields about a 1.5 mph (2.5 km/h) raising or lowering of the mean speed, respectively. When this statistic is combined with the power formula equating change in mean speed to crash risk, it is evident that lowering the speed limit will reduce crash risk, and raising the speed limit will increase crash risk.

Whether the safety gains/losses associated with the change in the speed limit is worthwhile must be examined in the context of maintaining reasonable mobility, and other system objectives. In addition, the policy context must be considered because the relationship between travel speed and speed limits indicates that the percentage of violators increases when speed limits are lowered and decreases when speed limits are increased.