| The most widely used model of the household transactions demand for money, the Baumol-Tobin inventory-theoretic model, is inadequate in several respects. This model provides a microfoundation for the relationship of money demand to income and interest rates, but does not provide a rationale for the empirical evidence which shows that money demand is apparently related to household wealth. Moreover, the model has not been updated to include recent research, which seems to show that credit used for purchases, the value of time, and time-varying non-monetary asset disposal costs are also important determinants of money demand. Finally, since the mid-seventies, the model has not forecasted well, generally overpredicting the money balances held by all economic sectors, including households. Together, these shortcomings point to the need for new thinking in this important area.; The capital-theoretic approach is an extension and generalization of the inventory-theoretic approach, based on the suggestions of Friedman and other money-capital theorists. Using microeconomic utility maximization and household production theory, the capital-theoretic model shows how household members use money balances as productive inputs, and thus how households make optimal investments in money balances based on productivity and cost principles. The model delineates the "purchase services" of household money balances, i.e. vis-a-vis credit as a means of purchase, money can often be used to purchase more optimal quantities of commodities, to purchase commodites at lower prices, to save commodity purchasing time, and to make unexpected commodity purchases at least cost. The model draws together the important features of the inventory-theoretic model and recent developments in money demand theory into one logically consistent model.; The capital-theoretic model adds several new variables to the conventional empirical money demand equation. Regression results generally verify the statistical significance of the new variables, but the results could be improved with the resolution of several data deficiencies. The model forecasts are marginally better than those of the conventional model.; The capital-theoretic approach is shown to apply to all balance-holding units, making it a truly general theory of the transactions demand for money. Further research is suggested. |
