Are measures of air-misses a useful guide to air transport safety policy?

Finding a proxy for analyzing and developing policy responses to rare events is often challenging. In the case of commercial airline accidents, these are small in number, especially within most national jurisdiction. This makes it difficult to use standard, objective probability techniques for assessing the impacts of various safety policies or for developing monetary measures for their inclusion in cost-benefit analysis studies. What is often used in place of the actual or predicted accident rate is a measure of ‘air’ (or ‘near’) misses that reflects the number of times a technically determined safety parameter is violated. The paper looks at whether this is a useful measure in terms of the quality of the data available, and whether it acts as a reasonable proxy when used in air transportation safety analysis and policy-making.