Skip to Content
drug information
Matters of Substance

High in the saddle: Cannabis-affected driving

Wednesday, August 20, 2008

New Zealand’s response to an apparent increase in drivers under the influence of cannabis seems likely to be more legislation and new penalties. But is this good policy, based on solid evidence? It might be, but Geoff Noller argues that the issues and evidence are complex and require much more discussion before we act.


In October 2007, Otago University published research suggesting driving under the influence of cannabis (DUIC) may be more common and riskier than driving drunk. The release of this research coincided with the introduction of the Land Transport Amendment Bill (No. 4), which creates a new offence of driving while impaired by illegal drugs. Intuitively, we might consider this an example of evidence-informed policy. It seems reasonable that cannabis use impairs driving, and there is evidence of DUIC in New Zealand, hence legislation.

As with so much about cannabis, however, the actual situation is complicated, giving rise to a number of questions.

What does the evidence say about how cannabis affects driving? To what extent is it a problem in New Zealand? If it is a problem, what response would be most effective?

Attempting to answer the first question immediately reveals the issue’s complexity. There is vast literature on the negative consequences of drunk driving, but there is much less on driving under the influence of drugs in general (DUID). What exists is contested, particularly where cannabis is concerned.

DUID may be examined through epidemiological analyses or by laboratory and experimental means where impairment is assessed by assigning people to either drug taking or placebo/non-drug groups and having them perform certain tasks. With driving, these include cognitive tests (i.e. divided attention), coordination and testing reaction time.

One criticism of this type of study is that laboratory tests do not necessarily reflect actual driving conditions or drivers’ experience. While cannabis studies do find deficits in concentration, reaction time, spatial and temporal judgement and attention to peripheral stimuli, these effects are less marked in non-experimental settings, and more experienced cannabis users appear even less affected. Other studies suggest that, while cannabis does impair driving,users are more aware of their impairment and respond by driving slower and less aggressively. This isn’t usually the case with alcohol.

Some researchers claim that more accurate data may be gathered via epidemiological studies such as the one by Otago University, which focus on DUID prevalence amongst subpopulations. These studies aim to describe the magnitude of the problem or analyse it in terms of which drugs are over-represented in accidents.

Two ways of doing this are through ‘case-control’ studies, where injured and non-injured drivers are matched and compared for their drug use, and through ‘culpability’ studies, where injured or killed drivers are assigned blame, relative to their use or otherwise of different substances.

But these studies also leave the cannabis case unresolved. One review article indicated a range across studies suggesting 2.7–13.9 percent of drivers had used cannabis. However, the authors noted it is difficult to determine causality, as the relationship between injury or death and drug use is one of association.

Also, the psychoactive metabolite of cannabis degrades relatively rapidly, potentially before it can be identified,which might lead to underestimated impairment. Alternatively, the inactive parent molecule remains longer, potentially providing a false positive for impairment.

Adding to these difficulties is some curious data suggesting cannabis affected drivers are either at no greater risk or in fact are at less risk of an injury or fatal accident. This was the case in a large Australian study of 2,279 nonfatally injured drivers, which found cannabis-only subjects to be marginally less culpable than drug-free drivers.

In summarising the mixed data on cannabis-affected driving, a Canadian review noted that, while cannabis is the most commonly used illicit drug and there is evidence of it moderately impairing driver performance, epidemiological data fails to show it is a major contributor to traffic crashes.

Given New Zealand’s high prevalence of cannabis use, as well as research suggesting a high rate of intoxication while driving, it seems likely that DUIC is a significant issue.

Data from Massey University’s National Drug Survey 2001 noted an increasing trend to DUIC, particularly by youth. From 1998 to 2001, reported rates for those aged 15–17 years increased from 18 percent to 39 percent. This data supports the notion that, in some New Zealand subpopulations, DUIC is an accepted aspect of culture, but also that injury and possibly death may result, particularly where drivers are inexperienced or prone to risk taking.

How best to address these issues? An obvious response is legislation such as that currently before Parliament. Nonetheless, questions remain about its specifics. With cannabis, is impairment, or simply evidence of past use, to be acted on? How accurate are tests, and will particular populations, such as Mäori, be targeted?

Internationally, legislative responses to such questions have varied. In Sweden, there is zero tolerance for anyone identified as DUID, while Germany has instigated an upper limit for cannabis, similar to the commonly operating blood alcohol content (BAC). Supporters of zero tolerance claim any level is too high, given some evidence of impairment, and that, regardless, cannabis is illegal. However, the issue’s complexity implies that simplistic solutions may not provide the best outcomes.

In 2005, an international panel, having examined 140 studies, noted the advantages of having a BAC equivalent for cannabis, i.e. a BCC. With the evidence indicating that low levels of cannabis do not significantly elevate risk of accident and that drivers might test positive while not being impaired, zero tolerance and the potential consequence of criminalisation were described as counterproductive.

An interesting US study further supports arguments favouring a BCC. Researchers found that, where alcohol was more expensive and harder to obtain, and cannabis less interdicted (i.e. decriminalised for medical purposes), those aged between 18–24 would substitute cannabis for alcohol and that, in these circumstances, driver injury and death would reduce. In the reverse situation, with cannabis use more heavily penalised and alcohol seen as more attractive, driver injuries and fatalities increased. They concluded that, despite a potential increase of cannabis use resulting from decreased penalties, a net harm reduction would result.

This article has aimed to extend the conversation on cannabis into an area of recognised concern and one requiring practical response. The complexity of issues surrounding DUIC requires consideration based on evidence but also discussion. Though one legislative response might suggest zero tolerance for DUIC, other measures such as a maximum BCC combined with education may ultimately produce better long-term outcomes.

There are, however, barriers to implementing alternative policies. For example, education around the safe use of cannabis in a prohibition environment may be too difficult a nettle to grasp. Nonetheless, if we are unable to discuss these issues, we are likely to develop policies that, at best, have little impact on the problem or even make it worse.

A harm minimisation approach would suggest those with whom this conversation must occur are those most able to reduce the negative consequences of cannabis-affected driving – cannabis users themselves.

  • Geoff Noller is a Dunedin-based researcher. He recently completed a PhD with Otago University’s Department of Psychological Medicine entitled Cannabis use in New Zealand: perceptions of use, users and policy.


As well there could be bias in Candors retort .

There is so much conflicting evidence re effects of cannabis and driving one can extrapolate out whatever stance suits agendas.

If cannabis and driving is not addressed based on sound data , and wise choice ,rather than be a fix for what could well be a minor problem, a greater one will be created .

One needs to be aware there is little advocacy or representation in support of any cannabis related matters . Be it research or the like , it does not attract funding from vested interest groups.

But there is a strong funding lobby from organizations that do have a vested interest in diminishing the part alcohol plays and fictitiously enhancing that of cannabis.

The effort and results to ensure alcohol and pharmaceuticals were extracted from this latests DUID initiative is a prime example.

Lets talk about cannabis ..
But first let it be on a level playing field and compare apples with apples..


There is some bias in this presentation. Unsurprising maybe from an author who's worked with NORML. To poo hoo or imply education isn't worth it under Prohibition denies the work of the needle exchange. Several errors of omission here could mislead, but the major problem is advocacy of the idea that experts still think cannabis role in crashes is unresolved or not major.

An "assessment" supported by already yellowing reviews, and reference to a superceded Oz study of 2,279 drivers, which reported low crash culpability for cannabis users. The scientist involved in that prehistoric research (Drummer) was long past forced, by Cannabis Reformers constant regurgitation of his debunked work, to publicly renounce it's discredited results on National TV. Drummers' more recent research (2004)studied 3,398 fatally injured drivers using better drug testing methods. It found an average of 270% higher culpability where cannabis was used, and culpability that was 660% higher than sobre drivers if over 5ng THC(recent use).

It is incorrect to suggest addressing drug driving could increase alcohol harm. The strong emphasis on drink driving prevention has increased teen drug driving death rates in NZ, keeping teen deaths high. MADD (Canada) acknowledges DUID as a greater risk for teens than drink driving. It's also undermining of road safety workers efforts, to go repeating some crackpots' easily refuted musings that a liberalisation of pot laws, combined with poor liquor availability, may have reduced a road toll somewhere.

Increased pot use would not reduce tolls IF previous outpriced drinkers had substituted pot, as THC also raises crash risk. If this med-pot "Shangri La" lands' toll dropped, that was surely to do with the reduced alcohol consumption or unrelated factors. Certainly not to do with some magic pot driving cure, as med-users shared newly legal fortune cookies. Such theories are typical of the wishy washy theories spread by the less responsible cannabis crusaders. And set the stage for not infrequent multiple death and injury crashes caused by designated pot drivers.

Our members dispute that NZ should be wary of criminalising people who are truly guilty of drug impaired driving - if PC is to rule then lets unprosecute all drink drivers! The issue, as noted, is that we must be as sure they are guilty as with alcohol. We agree a zero THC tolerance law doesn't achieve that aim. Best practise, as suggested, would be a BCC for sole users that links to reasonable impairment/risk, accompanied by education.

But BCC and blood alcohol limit should be waived as meaningless where poly-drugging features. Cannabis users have higher rates of recent alcohol and other depressant drug use, and 40x the rate of recent designer drug use of non users (Oz Household Survey). Yes, the salvation of this needed law lies in getting the detail right. That requires conversations not only with users, but also DUID specialists, and the victims - as it is continues through the legislative process.

66% of drug related mortality among problem users is indirect eg by car crash. If our Government doesn't act DECISIVELY to protect other drug users and their mounting victims, on road where it counts, we can only assume, that it's as Darke et al worry; that the "reduction of crime and disease occasioned by illicit drug users deaths may be seen as a net gain to society."