This paper describes an investigation into the effects of different types of road improvement scheme on accident frequencies in rural roads throughout the UK. Changes in accidents at 2000 treated sites are compared with those observed at neighbouring sites and untreated sites. The study originally aimed to determine whether there was actually `accident migration', the offset of accident reductions, associated with treatment, by accident increases elsewhere. Using a comparison of the levels of apparent migration at treated and untreated sites, no evidence was found for a real migration effect. However, the results also provided information about the safety effects of road improvement schemes and traffic engineering schemes in general. They suggest that, on the whole, highway improvement schemes and maintenance programmes yield a modest benefit: an overall injury accident reduction of order 10% for treated sites. No significant reduction was found for any of the individual categories of treatment, which included: (1) surface maintenance; (2) traffic management; (3) junction and link improvements; and (4) schemes aimed specifically at reducing accidents. Possibly greater changes in fatal and serious accidents are still being analysed.
Abstract