The paper by Graham et al.  discusses a possible model to prevent drinking and driving by applying principles based on successful enforcement interventions and the support of research. The focus of the paper seems to be English-speaking countries, where the measures to reduce intoxication among bar patrons are more advanced than in other developing countries. I comment on how societies such as Brazil could deal with the problem.
Drinking locations have been presented as environments that could lead to high-risk behaviours, such as driving while impaired or violence [2, 3]. However, in Brazil, where alcohol is associated with 21% of car accidents , service to intoxicated people is still considered as a non-criminal offence, and a minor infraction. It is generally not enforced. The state and federal governments have attempted to reduce the numbers of injuries caused by drinking and driving in recent years by lowering the allowed blood alcohol content (BAC) for drivers (the level is currently 0.02). Some significant results have been achieved in a few places  where the law was enforced, but this has not been a reality in most states of the country.
The authors seem to defend the construction of a prevention model based mainly on governments as regulators and guardians of public health interests. Additionally, as the paper indicates, there are advocacy groups that can drive political will over the subject and play an important role by working together with governments to improve collective wellbeing. Nevertheless, it is important to stress that there are some groups whose commitment to the political agenda of reducing alcohol-related problems is doubtful.
In the last decade, the alcohol industry has been an active player on ‘responsible consumption’ issues, especially through their social aspects organizations. The popularization of corporate social responsibility (CSR) has led the alcohol companies to develop practices (and gradually improve them) on drinking and driving that include servers' training programmes. In Brazil, since the 2000s companies such as Diageo and AmBev have been sponsoring and developing activities to the hospitality sector (as part of their sales training programmes) that provide some information on alcohol intoxication . There is insufficient data to evaluate the effectiveness of those programmes in preventing driving while impaired, but alcohol companies in Brazil seem to introduce programmes in areas where government action is inefficient or absent.
In fact, several authors have pointed out that CSR cannot be understood without considering government actions [7, 8]. First, areas in which there is more public investment are less likely to receive voluntary efforts from alcohol companies. Secondly, CSR is an attempt to address the deficiencies and limitations of governments and international institutions in solving public alcohol problems caused by corporate externalities. Although the idea of private companies responding to social pressures may seem legitimate, the underlying motivation for CSR can be tricky . In the case of alcohol, there is evidence that the strategy might be solely protecting company interests rather than serving the public good [10, 11].
Finally, as the authors argue, the construction of a model to reduce intoxication among bar patrons should have clear and unbiased standards. This result can only be achieved by using democratic tools guaranteed by governments and supported by civil society, without the interference of private interests engaging in promoting their own interests.
Declaration of interests
The author is supported by a grant from the International Development Research Centre, and she would like to thank the São Paulo Research Foundation (process nº 2011/18963-4).