The effect of an ageing society on road safety : future population characteristics. Since elderly people show an increased injury risk an increase in the number of traffic related injuries is expected as society becomes older. The effect of an ageing society, which coincides with other trends, however, is not straightforward. Exposure and risk differ a lot between different age groups. Even within age groups important differences exist, often related to the socio-economic status. Changes in population composition (both size and share of the different groups) will therefore influence the number of traffic related injuries and the share of the different population groups in the number of injuries. An adequate planning on the short and long term is required to adapt to a changing road safety. Considering the complex relations between the relevant factors the use of a model is favored. The activity based model FEATHERS, developed within IMOB, offers these possibilities. The model allows to calculate the activity and travel patterns of the population as a function of several person and household characteristics. For this a synthetic population (a fictive population that resembles the actual (or expected) population) has to be built. The evolution of the number of traffic related injuries can be determined based on the evolution of its 3 components: exposure, accident risk and injury severity. Differences in exposure and its evolution are linked to the activity patterns. Furthermore the elderly are inclined to compensate for their limitations by adapting their travel patterns so to make sure the task difficulty is reduced. Functional limitations and the attitude and resulting behaviour of road users determine the accident risk. Injury severity comes together with the increased vulnerability of older persons in traffic. Time (and therefore evolution) has 3 aspects. The age effect describes the differences that exist between the different age groups. The mere changing of the share of the age groups will lead to a change in the number of road victims. A second aspect is the cohorte effect. Successive generations differ from each other in habits and behaviour. The current elderly behave differently from the former elderly and their successors will behave differently on their turn. The succession of the generations will therefore cause changes too. The third aspect is the period effect. The spirit of the times, the measures that are in place in a certain year, will influence the number of road victims in a way. To study the influence of an ageing society on the number of traffic related injuries a clear picture of the future population is required. Some general methods of research on the future are mentioned, followed by the different population projection methods. Forecasting populations can be achieved by a modeling approach in which the mechanisms of population formation is modeled. A more mathematical approach is found in the technique of synthetic populations. Society doesn’t only change in terms of age and gender. Some other socio-demographic and mobility related variables will evolve also. Since these will influence the number of road victims too (often by its component exposure) it is important to estimate these changes as good as possible. Among these variables are work status, level of education, possession of a drivers’ license and the number of cars owned by a household. For the level of education and work status existing projections are used to model these changes. For possession of a drivers’ license and the number of cars owned by a household models are built that take into account the three aspects of time. They also contain a spatial component. Especially in the large cities the probability of owning a drivers’ license or a car is significantly lower. A procedure to obtain a future population based on the available data is proposed. A synthetic population is built where the existing population forecasts are used as boundary conditions (the variables age, gender and municipality are controlled for). Changes in the socio-economic variables are applied in the sample used to construct the synthetic population. Starting from the base year sample the variables are adapted in a consecutive way. In each step transition probabilities are derived from the distributions found/predicted in the current and future populations. By applying a random generator it is determined whether the change actually occurs. The derived population is used in the next stage of this research as an input for the activity based microsimulation model FEATHERS. Exposure is estimated on an disaggregated level as a function of age, gender, mode choice and road type. The exposure will then be linked to the evolution found for the risk (at the same level of disaggregation). This will lead to an estimation of the number of traffic related injuries. (Author/publisher)
Abstract