Abstract Conventional wisdom dictates that the more we know about a problem domain the easier it is to predict the effects of policies in that domain. Strangely, this wisdom is not sanctioned by formal analysis, when the notions of “knowledge” and “policy” are given concrete definitions...