A 1982 Highway Safety Research Center study (1) of factors associated with driving while impaired (DWI) convictions indicated that blood alcohol concentration (BAC), represented by mg. /100 ml., and the number of previous DWI convictions were most strongly related to DWI conviction. After adjusting for the effects of these, race and sex were still found to be strongly associated with the case disposition of persons arrested for DWI. For example, females with no previous DWI convictions had a lower probability of being convicted than males, while those with two or more DWI convictions were more likely to be convicted.
In 1983 major revisions were made in North Carolina's DWI laws. An intent of these changes was to remove biases in adjudication by the elimination of plea bargaining. This study replicates the earlier one and examines the relationship of BAC, number of previous DWI convictions, number of DWI arrests in the previous two years, year of most recent conviction, and age, race and sex of the subject to likelihood of conviction. Findings are examined to determine the effectiveness of law changes in reducing the influence of demographic variables on DWI conviction patterns.