A RECENT study has attempted to provide a critical evaluation of some of the main savings, or consumption, functions proposed for the United States and has assessed the predictive accuracy of different forms of seven of these functions as applied to the early postWorld War II years, I947 to the first half of I950.' The results of this study are subject to the limitation of being based, in the main, on predictions made over a relatively short period of time and one in which more usual peacetime consumption-savings patterns may have been distorted by after-effects of the war, e.g., shortages of consumer durables. It is the purpose of this paper to test the validity of these earlier results in the light of the more recent trends in consumption and savings. In view of this objective, we shall begin by outlining the plan of the empirical work in the earlier study and summarizing the results obtained. Section II, the bulk of this paper, is devoted to an extension of the empirical analysis in the earlier study to I950-53; there is an appendix at the end of the paper providing additional details on particular questions. In section III there is a comparison of the results obtained for this later period with those for the earlier years, and a new set of conclusions is propounded.

HOW stable are consumer expectations? In view of the growing awareness of a relationship between expectations and consumer buying, and the concomitant rising popularity of intentions surveys and attitudinal surveys as indicators of consumer purchases, this question assumes a new importance. To what extent, for example, can attitudes and expectations ascertained at one moment of time be extrapolated six months or a year in advance, as is implicitly done when information of this type is used either alone or in conjunction with other material to predict the future course of expenditures? ' Basic to the use of attitude and expectations data in explaining and predicting human behavior and basic to the design of surveys for obtaining such data is the question of the length of time these data retain their validity. Some light is thrown on this question by data collected during an experimental monthly consumer panel operation conducted by the writer with the aid of the Bureau of Economic and Business Research of the University of Illinois. The results of an analysis of these data as they relate to this question are presented here.