The considerations discussed are: (1) the relative importance of perceived accident risk, levels of police enforcement and the legal limit in determining BAC levels among drink drivers, and the quantitative effects of changes in these variables; (2) the increase in accident risk with BAC for different road user groups and injury severities; and (3) changes in the numbers of drink drivers detected and prosecuted at different BAC levels. The reason that each of these factors is so important is that accident risk increases very rapidly at higher BACs, especially for fatal accidents, whereas a policy of lowering the legal limit would bear mainly on drivers at the lower end of the alcohol distribution where any effects on accident risk would be relatively small. What makes the analysis so interesting is that the statistical distribution of driver BACs provide no indication of the legal limit in force, which is in marked contrast to the clear influence of speed limits on the distribution of vehicle speeds. This paper sets out the requisite statistical models for each of these factors, based on earlier research, the most recent available UK data and other countries' experiences of drink drive legislation. The consequences of lowering the drink drive limit to 50 mg/100 ml are then considered in the light of a range of enforcement policies; the levels of accident reductions and police resource implications are quantified and the likely importance of drink drive policy in setting future UK accident reduction targets are presented. The paper concludes with a discussion of the extension of the analysis to the zero limit case which highlights the social issues involved in setting acceptable levels of risk.
Abstract