The goal set up in Sweden is that the number of traffic fatalities shall not exceed 400 in the year 2000 and 250-270 in 2007. In order to monitor the attainment of this goal, a prediction model based on historical data on the number of fatalities and traffic is used. The model is simple, and only assumptions regarding traffic development are needed for future predictions. The model is updated every year. This study shows that the prediction model has so far been very stable. In addition, observed and predicted values have been in very good agreement. For the years 1994-96, however, the actual number of fatalities was relatively low. According to the study, this can to a large extent be explained by fewer young people and less alcohol in traffic. On the other hand, 1997 and 1998 may be characterised as less good from the stand-point of traffic safety. On the assumption that future increase in traffic will be ca 1% annually, it should at any rate be possible to get near the figure of 450 fatalities in 2000 and 350 in 2007 - provided that people who have actually died as a result of illness, ca 30-40 annually, are excluded. Should, however, traffic increase by ca 3% or more in both 1999 and 2000, it is not likely that the number of fatalities in 2000 will be much less than 500. (A)
Samenvatting