Older drivers have elevated crash rates in many countries, which can be used as a justification of specific licensing conditions targeted to this group, including on-road driving tests. In New Zealand, an older driver on-road driving test was introduced in 1999, which, together with certification of medical fitness, was a biennial requisite of driver licensing for drivers aged 80 and over. Following a medical check that health and eyesight were satisfactory for driving, the older driver undertook an on-road test of about 20 minutes’ duration in which their ability to detect traffic hazards around the vehicle, as well as their ability to control the vehicle and adhere to the road rules, was assessed. The driving test was failed if the driver committed major errors (e.g., failure to stop or maintain vehicle position in the lane) or a combination of more minor errors.
This test has been shown to provide some predictive value for crash involvement: a study showed that failures in the driving test were associated with higher crash involvement rates, controlling for relevant other factors such as driver age . Nevertheless, the generalised use of the test was dispensed with in December 2006, a decision supported by local and international evidence that, per licensed driver, older drivers on average do not have elevated crash involvement rates, particularly when taking into account their propensity to be injured when crash-involved [2–5].
From 2007 onwards, drivers at age 75, 80 and every two years after that were still required to obtain a medical certificate from their general practitioner (GP), who could make the following recommendations: the patient is medically fit to drive without conditions imposed; the patient is medically fit to drive with specified conditions (such as no night driving; only driving within 10 km of their home); the patient is medically fit to drive but must undergo an on-road driving test; the patient requires further specialist assessment (a medical specialist or an occupational therapist) before they can be deemed medically fit to drive; the patient is not medically fit to drive. Previous analysis of older driver failure rates found that about 5% of drivers aged 80 plus who sought to be licensed failed the on-road driving test, often after more than one attempt . Under the new licensing system (from 2007 onwards), these drivers would presumably still be driving unless identified by their GP as being unsafe.
When it was compulsory, the on-road driving test was reportedly experienced as very stressful for older drivers, and was often cited as a reason for relinquishing the driving licence, leading to an impaired level of independent mobility for many . Driving cessation is strongly associated with decreased out-of-home activity after adjusting for sociodemographic and health-related factors . Depressive symptoms can arise in response to consequent reduced access to resources such as paid or voluntary work, health care services, social contacts, etc. . A six-year US study found that driving cessation was the strongest predictor of depression among a sample of older people after adjusting for sociodemographic and health-related factors . Relatedly, Metz  found that a lower level of mobility was associated with a decrease in quality of life, particularly when drivers were already affected by age-associated disability. Marottoli et al.  hypothesised that for people with mobility issues, a sense of control over their environment is severely impaired by the lack of ability or opportunity to drive.
Removing the test requirement for older people might cause a fall in active travel (particularly walking) as a proportion of mode share, but time spent walking could feasibly increase due to higher levels of out-of-home activity . For example, upon driving cessation, older people who previously drove to undertake shopping may no longer do their own shopping, thus relinquishing walking activity that would have taken place to and from their car and around the grocery store. Nevertheless, an analysis of New Zealand Travel Survey data for people aged 75 plus who had ceased to have access to a car found that walking and passenger trips grew on average as a proportion of mode share, but mobility overall decreased . It would therefore be expected that the removal of the compulsory on-road driving test would have two main consequences for older people: improved access to motor vehicles; and consequent higher crash rates. The higher crash rates would arise from an increase in the amount of driving undertaken by this group, but could potentially be elevated further by allowing higher risk drivers to remain licensed (who would otherwise have failed the on-road licensing test). Some previous studies have compared older driver safety in jurisdictions with different older driver licensing policy [12, 13], finding little safety benefit from stricter regulation. However, no studies to our knowledge have looked at changes occurring in the same jurisdiction before and after a significant policy change. The current study aimed to identify any changes in licensing rates, crash rates, amount of driving and pedestrian activity by the group affected (New Zealand residents aged 80 plus) in comparison with other age groups, associated with the on-road driving test licensing requirement over the period 1999 to 2006.