Professor Friedman's Consumption Function and the Theory of Choice
W. M. Gorman
Turning now to the Friedman theory of consumption, there is a basic assumption which is fundamental to this theory which must be challenged. This is that the utility function can be assumed to be homogeneous in the value of consumption at different dates. For this to be true for all possible sets of prices for the different commodities it would be necessary for the utility function to be completely homogeneous in the quantities of all the goods consumed. And, if this were true, then a change in the level of income would produce proportionate changes in the consumption of all commodities. This is so contrary to common sense and the results of empirical research that the assumption of complete homogeneity must be dismissed. It follows that Friedman's basic assumption of the homogeneity of the utility function considered as a function of the total value of consumption at different periods could only hold true for certain special sets of prices. And it seems a priori rather unlikely that one of these special sets of prices should happen to coincide with the actual and expected market prices.