"Learning" is broader and more complex than simply the orderly acquisition of new knowledge. At least as important is the evolution of the background of assumptions and beliefs held by the community, or its principal decision makers, and implicit in its institutions and policies. These may bear only a loose relation to evidence or knowledge narrowly defined. The pressures of cost escalation over the past twenty years, and the attempts at containment in the U.S. and Canada, have added substantially to our knowledge of how the health care system works. Containment is possible, and the successful mechanisms, thus far, are quite specific. But the results of these attempts and (in the U.S.) the continued escalation have also significantly shifted the broader set of assumptions in the community about appropriate priorities and policies in health care. Attitudes towards physician supply, variations in practice patterns, capitated practice, and for-profit organization, for example, have changed radically, although the supporting evidence has not. But cost pressures have created an audience which wants to hear, whose background assumptions provide a different "fit" for the evidence.