THE CONSUMPTION FUNCTION
It is not the purpose of this book, in the present chapter or any other, to judge either the validity or the stability of any functional relationship. Prevalent opinions are mentioned, but they are not necessarily correct. Rather, our objective is to provide some insight into the mechanics of functional relationships, together with illustrations of a few that are more commonly recognized. It is the duty of every economist who uses them in analytical studies to decide for himself whether the functions he depends on are conceptually valid in general and appropriate to his particular objectives.
Functional relationships have become the core of national income analysis. Unlike the accounting relationships, however, they are not readily measured. While many economists have attempted to quantify them, no one can be certain of his measurement, and no one knows when they change. Functional relationships represent the willingness of persons or groups to take subsequent action based on flows that are current or past, or on expected future events. The economic accounts, however, disclose only accounting relationships, or the simultaneous occurrence of variables such as consumption and income.
Suppose an economist, presented with the national income accounts for a given day, say, Friday, should notice that personal disposable income is $100 and consumption $80. Should he conclude that households normally consume 80% of their income, or that some recent changes have set in motion multipliers that are not yet fully worked out, and the usual proportion of spending to income is something different? It might be, for example, that on Thursday disposable income was $80 and consumption $80, or 100% of income. On Friday, let us suppose that there was an exogenous increase in output (e.g., for investment, exports, or government consumption) of $20, and that this event brought income up from $80 to $100. Thursday's accounts may have been "normal"--that is, the