Limitations of risk approaches

ABSTRACT Risk methods are powerful and versatile, but they have limitations and subtle traps. The paper explores the nature of risk. There are three main difficulties. First, there is a problem with quality and completeness of information. Lack of precise information means that likelihood and consequences have to be estimated, so the information is vulnerable to biases: some are explored. Secondly, there is a serious problem of completeness in risk models, where omissions can lead to serious consequences. Unexpected events, sometimes called ‘black swans’, abound. Thirdly, the conjunction of very small probabilities and major consequences can lead to unreliable and dubious results. Applications areas considered are structural engineering, project management and risk management generally. There are situations where a resilience approach is preferable to risk.