Advertising and the Aggregate Consumption Function

The economic effects of advertising have been a much studied and hotly debated topic for a number of years. By now, there is fairly general agreement that, inter alia, advertising is important as a barrierto-entry (see Joe Bain, William Comanor and Thomas A. Wilson, Leonard Weiss) and that advertising does succeed in shifting demand for individual products (see Neil Borden, Nicholas Kaldor, Robert Dorfman and Peter Steiner, Lester Telser (1962), Kristian Palda), but there is little agreement as to the effect of advertising on aggregate consumption. John Kenneth Galbraith would have us believe that much of consumers' spending is managed from Madison Avenue,' but such a view has still to find universal acceptance.2 What is surprising, however, is that no one who has been party to the rather spirited debate generated by the Tall Gentleman's thesis has seen fit to examine by econometric methods the proposition that advertising has an impact on the aggregate consumption function. To undertake this is the purpose of this paper. In a modest, yet not insignificant, way, we feel that we have made some progress. Based on an analysis of advertising expenditures in the aggregate, our results suggest that advertising does in fact tend to increase consumption at the expense of saving. But as to what the causal mechanism underlying this is, we unfortunately cannot say. It may be that advertising actually succeeds in altering tastes a la Galbraith, but then again it may be that advertising is simply serving to bring new goods and services to the attention of consumers. As already noted, our analysis concentrates on the effects of advertising in the aggregate, and is conducted in the framework of the state-adjustment model of Hendrik Houthakker and Lester Taylor, as applied to aggregate consumption. Following Houthakker and Taylor, two variants of the model have been employed; the first focuses on consumption, and the second on personal saving. Section I presents a brief description of the Houthakker-Taylor (H-T) model and discusses the ways that it can be extended to accommodate advertising. This section also provides a short description of the data and methods of estimation. Sections II and III are empirical, Section II being devoted to a presentation of results and Section III to their critical evaluation. The paper is then concluded with some final observations in Section IV.