Abstract
We explore the interface between operational research, social sciences and risk research. The paper examines why different methods/tools produce different results/decisions, and what properties (characteristics) are embedded in the methods/models themselves. Emphasis will also be given to the failure, rather than success, of sensitivity and scenario analysis dealing with uncertainties. Deterministic models fail to take into account the value of flexibility. We shall show that solutions stemming from these methods may lead to trade-offs between cost cutting and risk, which are only a result of the chosen method, and not rooted in reality. Other solutions, coming from more suitable procedures, will lead to solutions, which are better in both respects: they cost less and are less risky. Psychological biases and organisational absorption of uncertainty may exacerbate a cleavage between safety and economy. Moreover, the value of flexibility is underestimated if companies apply too high discount rates or too pessimistic market forecasts. Failing to plan for an uncertain future may force actors in the future to improvise, and at the same time deprive them of the resources needed for improvisation.