Thresholds and Transmissions: Youth Drink-Driving and the Intergenerational Link Evidence from New Zealand

Abstract

Young people are disproportionately represented in drink-driving statistics, and a first conviction can have lasting economic and social consequences. In New Zealand, the Graduated Driver Licensing System (GDLS), the minimum legal purchasing age (MLPA), and the zero-alcohol limit for drivers under age 20 create three key legal thresholds at ages 16, 18, and 20 that shape young people’s access to driving and alcohol. This thesis uses linked administrative microdata from Statistics New Zealand’s Integrated Data Infrastructure (IDI), covering 427,899 young New Zealanders born between 1996 and 2003, to examine the causal effects of these thresholds on drink-driving offending and whether these effects differ by parental drink-driving conviction history.

Using a cohort-based parametric regression discontinuity design, I estimate discrete changes in offending at each threshold and examine heterogeneity by socioeconomic deprivation, geographic location, and ethnicity. The results show that turning 16 is associated with a gradual increase in drink-driving offending among males, reflecting increased access to driving. Turning 18 produces a sharp and persistent increase in male offending, consistent with legal access to alcohol. At age 20, offending appears to decline sharply, but this reflects reclassification of drink-driving offences as the legal blood alcohol limit increases, rather than a genuine reduction in risky behaviour.

Youth with parental drink-driving convictions exhibit substantially higher offending rates at all thresholds. However, their proportional response to each threshold is broadly similar to that of youth without parental convictions, indicating that the observed gap primarily reflects higher baseline offending rather than greater sensitivity to legal changes. This suggests that age-based policies alone are unlikely to eliminate intergenerational disparities. The highest risk is observed among youth in high-deprivation areas, rural areas with limited transport alternatives, and Māori youth, who are disproportionately represented among those with parental conviction histories. Although drink-driving rates have declined across successive birth cohorts, the absolute gap by parental background persists.