The publication of the paper has been delayed in order to have it peer reviewed and updated. It contains a critical analysis of the scientific evidence on the effect of reducing the limit and confirms that much of the research was flawed by failure to control for other measures being introduced at the same time and the existing downward trend in the statistics, lack of alcohol specific statistics and inappropriate interpretations of the evidence.

BAC limits have been used for more than 60 years to define impaired driving following the development of a relatively simple test to determine the amount of alcohol in a drivers body by Widmark in the 1930’s. This was first enshrined in per se laws in Norway in 1936, and Sweden in 1941, not be adopted elsewhere until 1967 in the UK with the British Road Safety Act, then many other countries followed suit. Presumptive limits of 100mg were used in the US as early as 1939. The breathalyser was developed by Professor Borkenstein in 1954. Limits are not enough however, conclude the authors - it is the perceived increase in the probability of being arrested and convicted for an impaired driving offence that is the most effective deterrent. ‘Per se laws did not prove a panacea for reducing impaired driving or reducing the alcohol-crash problem. In fact, in many jurisdictions (including Canada) the problem continued unabated. Hence, in the years following the introduction of per se laws, many jurisdictions began implementing other types of measures to support or enhance drinking-driving legislation. To a large extent, such measures concentrated on increasing the detection and apprehension of offenders — e.g., sobriety checkpoints, preliminary breath testing, and random breath testing. Administrative licence revocation has also proved to be a popular — and effective — measure. Recent efforts have also focused on programs for "hard core" offenders — e.g., mandatory assessment and treatment programs, vehicle impoundment, and alcohol ignition interlocks. Amidst the development and implementation of all these other programs, interest groups and legislators repeatedly return to the issue of the BAC limit, partly because of its apparent simplicity, straightforward rationale and perceived effectiveness. As a result, many jurisdictions have lowered, or are considering lowering, the per se BAC limit in an effort to further reduce the alcohol-crash problem.

The report notes that lowering the limit to 50mg in Canada would double the number of drivers liable for criminal prosecution. The situation would be similar in the UK where a dramatic increase in police resources would be needed to deal with these drivers, most of whom will show no visible behavioural signs that they are breaking the law. Without this increase many would go undetected and the deterrent effect of risk of being caught would be reduced.

The report then goes on to study in detail the effect that lowering the BAC limit has had in various US states and draws on comprehensive studies and research. It concludes ‘Every state that experienced a decrease in alcohol-involved fatalities had ALR in effect. In addition, two of the four states that introduced a lower BAC limit and ALR within a year of each other showed a significant positive effect when both laws were modelled as a single intervention. This would suggest that, by itself, a lower BAC limit had little impact on alcohol-involved fatalities At best, as many of the reports concluded, a lower BAC limit may have been but one factor that contributed to the continuing downward trend in alcohol-related crashes in the United States.’

The report then studies the effect of lowering BAC levels from 80mg to 50mg, as has been adopted by many EU members; Regarding the EU the report states that "it is perplexing that the push for harmonisation of BAC limits has focussed exclusively on the limit and not the sanctions associated with it. Pressure to adopt the same limit in so many countries where the penalty structure is profoundly different seems curious indeed from a traffic safety perspective." The report states that other actions with a similar risk as driving with a BAC of 50mg such as travelling at 65 km per hour in a 60km limit would not be subject to criminal prosecution or sanctions. (40 mph in a 37mile limit)

Great hopes and expectations have been placed on the potential of lower BAC limits. The fact that research has failed to demonstrate a strong, consistent effect raises the question, "Why?"

The rationale for lower BAC limits is predicated on the assumption that drinking drivers will comply with the new, lower limit by reducing the amount of alcohol they consume prior to driving, thereby lowering their risk of crash involvement. This, in turn, would lead to fewer alcohol-related crashes. There are, however, fundamental flaws in this logic. Knowledge of the law; understanding of the law in terms of one’s own behaviour; enforcement of the law; change in drinking and/or driving behaviour leading to a change in risk would lead ultimately, a reduction in crashes. ‘In many ways, lowering the BAC limit is a measure directed at the wrong group of drivers. To have an impact on crashes, a countermeasure must operate in such a way as to prevent drinking drivers who are most likely to crash, from driving after drinking — or at least, to prevent them from driving after consuming too much alcohol.’ The majority of drivers involved in fatal alcohol-related crashes typically have BACs in excess of 150 mg/dL, well in excess of existing BAC limits. The behaviour of this group of drivers has been the most difficult to change with any type of drinking-driving countermeasure. ‘There is certainly no reason to believe that the introduction of a lower BAC limit would somehow cause these drivers to obey the new limit when they fail to comply with the existing limit.’

The reports main conclusion, following detailed analysis of BAC law, its history and its enforcement around the world is that ‘what is important is to have a BAC limit and enforce it well.’ The actual numerical value of the limit may be of relatively little importance compared to the policies, programmes and procedures that have to be implemented to support it. There is little to be gained form changing it. ‘The evidence fails to provide a strong bases for supporting a public policy to lower the existing BAC limit. Our critical review of the research failed to provide strong, consistent and unqualified support for lowering the BAC limit for drivers in Canada. Therefore, it is our opinion that lowering the BAC limit from 80 mg/dL to 50 mg/dL would have little, if any, impact on the magnitude of the alcohol-crash problem in Canada.