Drink driving enforcement and publicity campaigns: are the policy recommendations sensitive to model specification?

Since their implementation, the drink driving enforcement and publicity campaigns in the Australian State of Victoria have been extensively evaluated and the mixed results obtained have generated much public debate. Using the same data from previous studies, this paper re-evaluated the effectiveness of the campaigns and tested several model assumptions and specifications. In general, the results obtained were robust and showed that the campaigns were effective in reducing serious crashes during high alcohol hours.