Which cost of alcohol? What should we compare it against?

This paper explores and develops issues raised by recent debates about the cost of alcohol to England and Wales. It advances two arguments. First, that the commonly used estimates for alcohol harm in England and Wales are outdated, not fully reliable and in need of revisiting. These estimates rely on data that are between 4 and 12 years out of date and sensitive to questionable assumptions and methodological judgements. Secondly, it argues that policymakers, academics and non-governmental organizations should be more careful in their use of these numbers. In particular, it is imperative that the numbers quoted fit the argument advanced. To help guide such appropriate usage, the different types of cost of alcohol are surveyed, alongside some thoughts on the questions they help us to answer and what they imply for policy. For example, comprehensive estimates of the total social cost of alcohol provide an indication of the scale of the problem, but have limited policy relevance. External cost estimates represent a 'lowest common denominator' approach acceptable to most, but require additional assumptions to guide action. Narrower perspectives, such as fiscal, economic or health costs, may be relevant in specific contexts. However, optimal policy should take a holistic view of all the relevant costs and benefits. Similarly, focusing solely on tangible costs may be less controversial, but will result in an under-estimate of the relevant costs of alcohol.