Evaluation of the Random Breath Testing Initiative in Victoria 1989-1991. Multivariate time series approach

Monash University Accident Research Centre - Report #38 - 1992

Authors: M. Cameron, A. Cavallo & G. Sullivan

Full report in .pdf format [3.5MB]

Abstract:

A quasi-experimental time series evaluation of the effect a random breath testing (RBT) initiative, introduced late in 1989 in Victoria, on crashes during 1990 was undertaken (Report no. 37). This report presents an additional evaluation study which uses an alternative method of estimating the effect of the initiative on crashes it) 1990, and also attempts to assess the effect during 1991.

The RBT initiative involved a substantially different method of RBT enforcement compared with past operations, with bus-based RBT stations replacing car-based stations and a multi-million dollar, Statewide anti-drink driving publicity campaign through all mass media.

Multivariate time series modelling of high alcohol hour serious casualty and fatal crashes was undertaken to estimate the change relating to the RBT initiative during 1990 and 1991, taking into account changes in unemployment rate and changes in the same crash types in NSW. A form of time series modelling known as ARIMA Intervention Analysis was used to estimate effects during 1990, whilst a multiple regression approach was used to estimate effects during 1991.

The findings of the present study indicate that the PBT initiative (in its entirety) resulted in an 18% reduction in high alcohol hour serious casualty crashes and a 24% reduction in high alcohol hour fatal crashes in metropolitan Melbourne in 1990, but no statistically significant effect during 1991. In rural Victoria, high alcohol hour serious casualty crashes decreased by 13% in 1990 and by 24% in 1991, whilst there were no statistically significant effects in high alcohol hour fatal crashes in rural Victoria in 1990 nor 1991. The conclusiveness of these findings depends on the adequacy of unemployment rate as an indicator of changes in travel during high alcohol hours, the appropriateness of NSW as a comparison area to take into account the effects of "other" factors (other than unemployment rate) influential in Victoria during the intervention period, and the assumption of minimal effects of concurrent speed camera operations in Victoria during high alcohol hours.

Sponsor: Transport Accident Commission