Author Sara Rahman
Published August 2022
Report Type Crime and Justice Bulletin No. CJB251
Subject Alcohol; Driving offences; Evaluation reports
Keywords Driving offences, Alcohol, Regression discontinuity, Difference-in-differences



The Mandatory Alcohol Interlock Program (MAIP) was introduced in February 2015 and applies to offenders convicted of refusing a breath test, high range drink driving and repeat drink driving. After serving an initial disqualification, offenders can choose to:
• drive with an interlock device in their vehicle, which requires a negative breath test to start the vehicle; or
• serve out the remainder of a 5 year automatic disqualification period.
To evaluate the impact of the program, we used a dataset of 98,501 proven drink driving and refuse to provide a breath sample offences committed between 1 February 2012 and 30 April 2018, linked to Transport for NSW datasets on MAIP orders and crashes. We compared outcomes for first-time offenders who took up MAIP and who recorded a Blood Alcohol Concentration (BAC) just above the 0.15 threshold, with outcomes for first-time offenders just below the threshold who were likely to have taken the program up were it available. To estimate the program’s overall impact, we also compared outcomes for offenders who committed MAIP-eligible offences with offenders who committed first time mid-, low- or special-range drink driving offences, before and after MAIP was introduced.

Key findings

Figure 1. Change in the probability of committing a drink driving re-offence before
 Figure 1. Change in the probability of committing a drink driving re-offence before
Figure 1 shows the change in the probability of committing a drink driving re-offence before, during, and after interlocks were installed in the vehicles of first-time high-range drink driving offenders. First-time offenders were 11 percentage points less likely to commit a new drink driving offence while the device was installed, than their counterparts who recorded a BAC slightly below the 0.15 threshold. We also find a small (3p.p.) reduction in the likelihood of a drink driving offence in the two years after the interlock is removed.
Looking at all eligible offenders, we estimate that MAIP reduced the likelihood of drink driving reoffending by 3.4 p.p. within 36 months of finalisation, and by 6.0 p.p. within 60 months of finalisation. Both of these constitute a 43% reduction in drink-driving reoffending among all serious and repeat drink driving offenders eligible for Phase 1 of MAIP.
We found no impact of the program on driving while disqualified or on road crashes, despite the program enabling offenders to return to driving sooner.


Interlock devices significantly reduce drink driving while they are installed and (to a modest extent) following their removal.