Income Redistribution and Aggregate Consumption: Implications of the Relative Income Model

James Duesenberry's classic model of consumer behavior (1949) neglects to show precisely what would happen in the model if a given aggregate income is redistributed. A version of Robert Frank's 1985 model of the demand for unobservable and nonpositional goods is used to establish a correspondence between the income and consumption distributions and to simulate the effect on consumption and savings rates of 10% and 50% redistributions toward the mean. The results show that redistributions unambiguously raise the average propensity to consume (APC) when concern for relative standing takes the form of concern about consumption rank. The APC, however, is only modestly affected by even the largest redistribution. This can be explained in part by a unique feature of the consumption function. When relative standing concerns are weighted heavily in the consumption function, the marginal propensity to consume falls off significantly but begins to rise at very high income levels.