The pandemic has highlighted how difficult it is to judge the wider welfare effects of responding to a major crisis. Strategies used to manage public health risks themselves pose risks: to those whose healthcare needs cannot be met due to the diversion of resources, to the economy, to the potential victims of domestic violence, to the education and welfare of children, to name only a few. These impacts are unequally distributed across countries and among groups within societies. How should policy makers and businesses make and evaluate decisions which involve trading off competing priorities, which cannot rely on simple assumptions of rational behaviour in anticipating how people and societies will respond, and in which all our behavioural biases will come to bear in deciding what to do, what data to use and whom to trust? How can we ensure that the tools we use to quantify (and ideally monetise) the full range of costs and benefits of an intervention, are appropriate? What is the role for technocratic experts in making those assessments, and to what extent should we rely on political processes to guide decision-making?