Assessing the Impact of Speed Limit Increases on Fatal Interstate Crashes By Sandy Balkin, Ernst & Young LLP and J. Keith Ord, Georgetown University
Executive Summary Introduction Data Source Methodology Key Points Charts and Tables Fig 1: Trend Component for Rural Arizona Fig 2: Seasonal Component for Rural Arizona Figure 3: Irregular Component for Rural Arizona Figure 4: Significance Level of First Rural Limit Change Figure 5: Significance Level of Second Rural Limit Change Figure 6: Significance Level of Urban Limit Change Figure 7: Monthly Fatal Crashes Figure 8: Rural Seasonal Component Significance Level Figure 9: Urban Seasonal Component Significance Level Table 1: Significant Rural Changes Table 2: Significant Urban Changes Table 3: Predicted number of fatal crashes attributed to the speed limit increase on rural interstates. Table 4: Predicted number of fatal crashes attributed to the speed limit increase on urban interstates.	Methodology The data used in this study are the number of fatal crashes for each month (from January 1975-December 1998) for each state separated by rural and urban interstates. Such data are known as a time series. Time Series Analysis is the field of studying, modeling and forecasting processes that change over time. When we consider the impact of an increase in speed limit on the number of fatal crashes, we are mainly concerned with the modeling aspect of time series analysis, looking backwards in time to see "what happened." There are many techniques for this type of analysis. Most of these techniques assume that, over time, the average level and variation of the series both stay constant. In practice, this is rarely the case. In 1986, Andrew Harvey and James Durbin produced a study entitled "The effects of seat belt legislation on British road casualties: a case study in structural time series modeling," which appeared in the Journal of the Royal Statistical Society–Series A, volume 149. In their investigation, these two outstanding researchers introduced a novel method of analyzing the impact of government legislation called Structural Time Series Modeling. Previous methods used for this type of study required the researcher to manipulate the series until its average level and variation remained constant over time. Thus, all obvious trends, cycles, seasonalities, etc. were removed prior to modeling. In the structural approach, models are set up explicitly in terms of components of interest such as trends, seasonals and cycles. In addition, instead of assuming that these components remain constant over time, this approach allows them to evolve through time. This approach is intuitively appealing since environments that generate time series often do not remain constant. The computer package STAMP© 5.0, developed by Harvey and his associates, was used to perform the analyses presented in this study. As an example of this method of analysis, consider rural interstates in Arizona. The speed limit was changed in April 1987 and December 1995. The original time series is shown in Figure 1. The series is decomposed into trend, seasonal and irregular components that are represented graphically in Figures 1-3. We can see visually that there is a significant increase in the trend around 1987 but none around 1995. This indicates that around 1987, the average number of fatal crashes significantly increased, but not so elsewhere. This increase occurs at the same time as a speed limit increase. Statistically, the 1987 speed limit increase resulted in a 41% increase in rural interstate crashes in Arizona. There is no statistical evidence that the 1995 speed limit increase had any additional effect on the number of fatal crashes. Next, we see from the seasonal component that there appears to be a strong monthly effect on the number of fatal crashes. For this example, there are considerably more crashes in June, July and August compared to the other months. Such seasonal patterns exist for most states and reflect the higher traffic levels in summer months. The irregular component is simply what is left over after the trend and seasonal components are taken into account. Since the irregular component is small compared to the actual series, most of the structure in the data is accounted for by the trend and seasonal components. The Structural Time Series Modeling approach tells us that there is a strong seasonal effect on the number of fatal crashes and that there is a significant increase in the number of such crashes around the time the speed limit was changed. However, we observe from the plot of the Trend component that it appears that after the initial jolt of the speed limit change, the trend gradually moves back towards the level prior to the speed limit increase. This phenomenon was observed for a number of states, but not for all. The overall U.S. figures in Figure 7 show a slight decline in the post-1987 period but the picture is less clear because the state laws were enacted at different times. There are several possible explanations for such reductions including: Drivers adjusted to driving at higher speeds States increased enforcement of driving laws Automobile safety was improved. However, we stress that our analysis was not designed to examine these questions; rather, they are important issues for further investigation. Next: Key Points -->