EDWARD GRAMLICH'S STUDY in this issue raises a question by a method that has frequently been found fruitful in scientific efforts. He presents an interpretation of observed data in terms of a widely used analytical framework that points to a very unfavorable distribution of the consequences of demand disinflation by undesirable output effects, on the one hand, and desired price effects on the other. Yet he directs attention to the possibility that a reconstruction of the standard framework could reduce the weaknesses he detects in the interpretation he presents.H e does not favor the specific reinterpretation that I have suggested in various writings and I will describe below-indeed, he leans perhaps more to an attitude of doubt in this regard-but he has an open mind about the methods of improving the conventional framework. Essentially,h e plays into the hands of all those of us who are even more skeptical than he is about the conventional way of going about macroeconomic modeling and who believe in the superiority of a specific alternative. I would like to make use of the opportunity his work provides. This is done partly with reference to his empirical findings, to which I turn in the third section, after an attempt to place the matter in perspective.