The six year before‐six year after ARIMA models used in this analysis demonstrate an increase in deaths and case‐fatality rates on all roads that was not only immediate and large, but was sustained. Prior ARIMA time‐series studies on the effects of speed limit increases have been restricted to short post‐intervention periods.20,21,22,23
These studies do not tell us whether the effects of raised speed limits on deaths, injuries and severity of injuries persisted years after the policy change.
The immediate increases in deaths and CFR predicted the subsequent sustained six‐year trend (figs 1–3). The largest increase in deaths occurred on interurban roads, but a spillover effect appears to have occurred on urban roads. The analysis indicates that 347 (10.6%) of the 3265 deaths that occurred between November 1993 and December 1999 were attributable to the increase in the legal speed limit, 230 (12.7%) of the 1817 deaths on interurban roads occurred as a result to the policy change, and 120 (8.3%) of the 1448 urban deaths.
This is the first study to evaluate the effects of changes in the legal speed limit on both deaths and case‐fatality rates using ARIMA time‐series models. Case fatality rate (CFR and CFRS) is a parameter independent of vehicle kilometres travelled and is a direct measure of the effects of speed of impact, seat belt use, and trauma care.24
The observed increase in deaths following the increase in the speed limit to 100 kph substantially underestimated the increase in deaths directly attributable to the increase in CFRS (Formula 1). The exponential relationship25
between “small” increases in speed and nationwide increases in CFRS over the next 6 years, more than offset the decreases in death risks from protective countermeasures such as newly widened roads, improved lighting, clover‐leafs, urban roundabouts, air bags, rear seat belt laws, more speed enforcement and trauma care, and other countermeasures, as well as more congestion.
Sporadic monitoring from 1971 to 1994 indicates that right after the change in speed limit policy on November 1993, travelling speeds increased on high‐speed roads by 4.5% on the slow lane of the Tel Aviv‐Haifa road, (Road 2) and by 9.1% on the fast lane of the Jerusalem‐Tel Aviv road, (Road 1) and even more so on a newly widened stretch of one major connecting road (highway #40)—all compared to the year before.16
Other data showed that speeds rose on all three highways after the speed limits were raised, and that the mean estimated increase in speeds on the high‐speed roads later fell back to some 4% (range: −4% to +13%) in 1995.26,27,28
The empirical relationships identified by Nilsson,29
following models derived from Newtonian physics, show that increases in travel speeds lead to increases in deaths to the fourth power. Based on Nilsson's models, we would have expected an increase of 17% in fatalities on interurban roads.29
We observed an increase of 12.7% (N
230) on interurban roads. If no countermeasures or increased congestion occurred during the post‐intervention periods we would have observed a 19.2% increase in fatalities on interurban roads (expected deaths divided by total injuries; 377/(1817+147)).
Theory suggests that serious injuries should increase following increases in travel speeds,24
but we observed an insignificant decrease in serious injuries. However, the predictive equations on the fourth power effect are based on the assumption that characteristics of the road environment remain relatively unchanged. In our case, there were massive changes in the road environment of high speed and all other interurban roads. The introduction nationwide of major improvements in road design—eliminating black spots, more night time lighting, road dividers between north and south bound traffic and road widening—may indeed have led to fewer crashes, but at higher speeds, so that case fatality and deaths greatly increased despite large falls in crashes and injuries
Some studies showed that the effects of increased speed limits are not limited to high‐speed roads. Speed adaptation and spillover effects occur when drivers coming off high‐speed roads continue to drive faster than those already on the same road13,30,31
and may account for more deaths than on the safer roads with higher design speeds. Therefore, studies in which time trends on lower speed spillover roads serve as controls for trends on high‐speed roads erroneously underestimate the effects of speed.16
The observed increase in modified CFRS and road deaths on urban roads provide strong evidence for a spillover effect.
Implications for prevention
The findings show that a small increase in the speed limit (6 mph) resulted in an immediate, substantial and persistent increase in road deaths and case‐fatality. In the US, approximately 42 000 Americans are killed each year in road crashes, yet US policy has been to end the national speed limit of 55 mph and increase travel speeds because there remains contention whether small increases in speed can result in persistent increases in road fatalities. The catastrophic failure of the US to reduce its road deaths in absolute numbers contrasts starkly with the trends in the UK, France and Australia, where there have been reductions of the 40–50% associated with the introduction of nationwide speed camera networks.32,33
More importantly, the persistent impact of increased speeds on road deaths was observed despite coinciding with tremendous improvements in road design, vehicle safety, mandatory seat belt and child restraint use, and other important safety countermeasures. These countermeasures do not prevent the increase in deaths associated with higher travel speeds.
- Prior autoregressive integrated moving average (ARIMA) time‐series studies on the effects of speed limit increases have been restricted to short post‐intervention periods.
- In the US, approximately 45 000 Americans are killed each year in road crashes, yet US policy has been to end the national speed limit of 55 mph and increase travel speeds because there remains contention whether small increases in speed can result in persistent increases in road fatalities.
- We use ARIMA time‐series models to show that a “small” increase in the speed limit (6 mph) resulted in an immediate, substantial and persistent increase in road deaths and case‐fatality six years after the change in policy.
- More importantly, the persistent impact of increased speeds on road deaths was observed despite coinciding with tremendous improvements in road design, vehicle safety, mandatory seat belt and child restraint use, and other important safety countermeasures. The opponents to stricter speed regulations often cite the improvements in safety and design as a justification for the increased speed limits.