Although most researchers working in the field, both in mainstream/neoclassical economics and in at least some heterodox approaches such as the Austrian School, would agree with the ideal of value freedom, among methodologists and philosophers of science the view that science in general, and economics in particular, cannot and should not be value-free is dominant. The aim of my talk is to add a new argument to the arsenal of the advocates of value-laden science. The argument is essentially this. Without theories about the nature of price formation and money, economics cannot get off the ground. However, different theories have different implications for which social phenomena and institutions should be considered "natural" and which should be considered "man-made", "planned" or "deviations from the normal". This in turn has implications for the desirability and justification of social interventions and thus for policy recommendations. In the choice between the basic theoretical models, evidence can only play a subordinate role. My main case study will be the controversy between "metallists" and "chartalists" about the nature and origin of money.