Optimism and calibration in risk assessment: a note in reply to Hartz and Elrod

Abstract

By way of a reply to Hartz and Elrod's discussion of optimism in risk assessment, we describe some findings of a study into the process of predicting the engineering resources associated with complex, technologically advanced systems in the aerospace industry. These suggest that biases in risk prediction stem both from limitations in individuals' cognitive abilities and from incentive effects. These incentive effects in turn arise partly from conditions in the organisation and partly from conditions in the organisation's environment. Our findings also suggest that incentive effects can lead both to under-estimation of future outcomes and over-estimation. We address the extent to which under-estimation, or optimism, is functional in the organisations in question.